design and development of a uav system for multispectral … · iii infrared bands. the cameras...
TRANSCRIPT
Design and Development of a UAV System forMultispectral Imaging and Remote Sensing
Applicationsby
Jorge Marin Marcano
A thesis submitted in partial fulfillment of the requirements for the degree of
Master of Science
in
Computer Engineering
Department of Electrical and Computer Engineering
University of Alberta
© Jorge Marin Marcano, 2018
ii
Abstract
The increasing use of the Unmanned Aerial Vehicles (UAVs) over the past few years has opened
the door to a large number of applications in research, commercial, and industrial areas. UAVs
have become an efficient, cost effective, and environmentally friendly (less or zero CO2 emissions)
platform for remote sensing compared to manned vehicles. One major field in this category is
multispectral sensing. Commercial solutions of UAV-compatible multispectral cameras are available
but tend to have a high cost and, in most cases, lack certain flexibility when it comes to custom
development. This thesis is organized in three basic sections. It begins with the development of
a modular fixed-wing UAV that includes a customized airframe, integrated avionics with a custom
UAV system board with redundant power supply, and single board computer (SBC) for payload
interfacing. The other two sections of this thesis are focused on the design, development, and testing
of two revisions of a custom multispectral imaging system.
The customized fixed-wing airframe was redesigned aiming for modularity and reliability. The
open-source UAV system board integrates all essential sensors and components in a small and com-
pact board, reducing the amount of wiring and connectors, which in turn reduces weight and in-
creases reliability. The UAV system board includes a detachable and novel redundant power supply
board that provides main and backup power regulation. The complete UAV system design is pre-
sented as an efficient, and flexible platform.
The multispectral imaging system is based on multiple USB 2.0 camera modules and optical filters.
The first iteration, or proof of concept, introduces a 3-band multispectral imaging system based on
3 rolling shutter cameras synchronized by software, with optical filters for the green, red, and near-
iii
infrared bands. The cameras were controlled by the SBC on board of the UAV. The imaging system
was tested over a grass field, and the images collected were post-processed. Resulting images were
aligned and NDVI images were produced where it was possible to detect vegetation and identify
different levels of vegetation stress. This system proved to be a low cost and simple, yet effective
approach for custom UAV multispectral imaging.
The next evolution of the system is a 12-band multispectral imaging platform based on USB 2.0
camera modules. Cameras used in this revision have a global shutter and a trigger functioning
mode that allows hardware synchronization. A novel bridge board was designed to integrate all
USB communications via USB HUBs and also include a microcontroller unit (MCU) in charge
of triggering the cameras. The USB HUB network has selectable routes for grouping cameras to
different upstream USB ports depending on the bandwidth requirements of the application. The
system was tested and characterized for the maximum number of simultaneous cameras per root
USB port depending on camera configuration. With the highly integrated bridge board the overall
size and weight of the system was reduced, making it compatible with most UAVs. Software was
also developed, based on third party tools, that allows extensive configuration of the cameras and
control over the imaging system.
The result of this work introduces a detailed methodology for UAV systems development, introduces
a comprehensive UAV system board, and explores the capacities of USB 2.0 camera modules as part
of complex imaging systems. The hardware and software platforms developed throughout this thesis
are cost effective and efficient solutions that allow customization and expansion, making them an
ideal development tool for UAV systems and UAV multispectral imaging applications.
iv
Preface
The work in this thesis, including the conceptualization, design, implementation, data processing,and analysis, was carried out by me under the supervision of Dr. Duncan Elliott from the Departmentof Electrical and Computer Engineering at the University of Alberta.
The work was done in working space provided by the University of Alberta Aerial Robotics Group(UAARG), and later in Dr. Michael Lipsett’s laboratory, from the Department of Mechanical Engi-neering at the University of Alberta.
v
Acknowledgements
I would first like to thank my supervisor Dr. Duncan Elliott from the Electrical and Computer En-gineering Department of the University of Alberta for his insights and guidance, and Dr. MichaelLipsett from the Mechanical Engineering Department at the University of Alberta for his support.
I would also like to thank the University of Alberta Aerial Robotics Group (UAARG) for their helpduring the first steps of the project. Many thanks to Andrew Jowsey, Sebastian Werner, Cindy Xiao,and Brian Hinrichsen. Their knowledge and experience were of great importance. I would like tospecially thank Rijesh Augustine for his vital contributions throughout the project. I would also liketo acknowledge Stream Technologies Inc. for their valuable support in the final stages of the project.
Finally, I must express my profound gratitude to my family, friends, and to my loving spouse, fortheir continuous support and encouragement throughout my years of study and through the processof conducting research and writing this thesis. This accomplishment would not have been possiblewithout them. Thank you.
Jorge Marin.
vi
Contents
List of Tables x
List of Figures xii
List of Abbreviations xvii
Chapter 1: Introduction 1
1.1 Objective of the Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Organization of the Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter 2: Literature Review 4
2.1 Imaging Systems and Cameras . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1.1 Fundamentals of Optics and Imaging Systems . . . . . . . . . . . . . . . . 4
2.1.2 Image Sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.2.1 CCD vs. CMOS Technologies . . . . . . . . . . . . . . . . . . 8
2.1.2.2 Rolling vs. Global Shutter . . . . . . . . . . . . . . . . . . . . . 9
2.1.2.3 Performance Metrics of Image Sensors . . . . . . . . . . . . . . 9
2.1.3 Lenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1.4 Modulation Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1.4.1 Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.4.2 Contrast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.4.3 Slanted-Edge MTF Measurement . . . . . . . . . . . . . . . . . 17
vii
2.1.5 USB Camera Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.1.5.1 USB Video Class (UVC) . . . . . . . . . . . . . . . . . . . . . 19
2.1.5.2 Video for Linux Two (V4L2) API . . . . . . . . . . . . . . . . . 20
2.1.5.3 GStreamer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.2 Spectral Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.1 Vegetation, Soil and Water Reflectance . . . . . . . . . . . . . . . . . . . 22
2.2.2 Landsat Thematic Mapper . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.2.3 Lightweight UAV-Compatible Multispectral Imaging Systems . . . . . . . 24
2.3 UAV-Based Spectral Remote Sensing . . . . . . . . . . . . . . . . . . . . . . . . 27
2.3.1 Crop Inspection and Precision Agriculture . . . . . . . . . . . . . . . . . . 27
2.3.2 Surface and Soil Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.3.3 Surface Hydrocarbon Detection . . . . . . . . . . . . . . . . . . . . . . . 30
2.3.4 Environmental Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.3.5 Archaeology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Chapter 3: UAV Development 33
3.1 UAV Electronic System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.1.1 UAV System Board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.1.1.1 Microcontroller Unit (MCU) . . . . . . . . . . . . . . . . . . . 35
3.1.1.2 Inertial Measurement Unit (IMU) . . . . . . . . . . . . . . . . . 35
3.1.1.3 GPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.1.1.4 XBee and Data-logger . . . . . . . . . . . . . . . . . . . . . . . 36
3.1.1.5 PWM Outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.1.1.6 Power Regulation . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.1.1.7 Additional Features . . . . . . . . . . . . . . . . . . . . . . . . 37
3.1.2 Redundant Power Supply Board . . . . . . . . . . . . . . . . . . . . . . . 38
3.1.2.1 Principle of Operation . . . . . . . . . . . . . . . . . . . . . . . 40
3.1.3 Firmware and Ground Control Station (GCS) . . . . . . . . . . . . . . . . 43
3.2 Payload Computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
viii
3.3 UAV Avionics Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.4 Airframe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.4.1 Falcon I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Chapter 4: Proof of Concept Multispectral Imaging System 53
4.1 USB Camera Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.1.1 Selection and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.1.2 Software Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.1.3 Lens Testing - Modulation Transfer Function . . . . . . . . . . . . . . . . 57
4.2 3-Band Multispectral Imaging System . . . . . . . . . . . . . . . . . . . . . . . . 58
4.2.1 Cameras and Lenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2.2 Optical Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2.3 Software Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.2.4 UAV Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.2.5 Flight Campaigns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.2.6 Image Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.2.7 Vegetation Indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Chapter 5: 12-Band Multispectral Imaging Platform 73
5.1 System Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.2 USB Camera Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.2.1 Optical Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.3 Camera Bridge Board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.3.1 USB HUB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.3.2 Trigger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.4 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.5 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.6 Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.7 Focusing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.8 Image Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
ix
Chapter 6: Conclusions and Future Work 91
6.1 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Bibliography 95
Appendix A: UAV System Board User Guide 105
Appendix B: Camera Bridge Board Schematics 154
Appendix C: Camera Bridge Board Bill-of-Materials 167
Appendix D: 3D Printer Settings 171
x
List of Tables
2.1 Features comparison between CCD and CMOS image sensors. . . . . . . . . . . . 9
2.2 Spatial frequency unit conversion chart. . . . . . . . . . . . . . . . . . . . . . . . 15
2.3 Landsat Thematic Mapper Bands for Vegetation Monitoring. . . . . . . . . . . . . 24
2.4 Comparison of common multispectral imaging systems. . . . . . . . . . . . . . . 26
3.1 UAV System Board Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2 Technical Specifications of the ODROID-C2 . . . . . . . . . . . . . . . . . . . . . 44
4.1 Features of the ELP-USB500W02M camera. . . . . . . . . . . . . . . . . . . . . 54
4.2 Settings of the ELP-USB500W02M camera. . . . . . . . . . . . . . . . . . . . . 55
4.3 Frame Rates for the ELP-USB500W02M camera based on output video format andresolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.4 Results of different M12 lenses tested using the slanted-edge MTF method. . . . . 57
4.5 Features of the ELP-USB130W01MT-B/W camera. . . . . . . . . . . . . . . . . 59
4.6 Settings of the ELP-USB130W01MT-B/W camera. . . . . . . . . . . . . . . . . . 60
4.7 Frame Rates for the ELP-USB130W01MT-B/W camera based on output video for-mat and resolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.8 Technical details of the optical filters. . . . . . . . . . . . . . . . . . . . . . . . . 62
4.9 Flight campaigns details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.1 Specifications of the DMM 42BUC03-ML camera. . . . . . . . . . . . . . . . . . 74
5.2 Video formats and frame rates available in the DMM 42BUC03-ML camera. . . . 76
xi
5.3 Settings of the DMM 42BUC03-ML camera. . . . . . . . . . . . . . . . . . . . . 76
5.4 Test results of USB traffic with different configurations. Cameras were connectedto a single USB upstream port and triggered simultaneously. Data transferred basedon a single frame from each camera. . . . . . . . . . . . . . . . . . . . . . . . . . 82
5.5 Comparison of common multispectral imaging systems with the system presentedin this work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.6 General costs of components for the multispectral imaging system hardware. . . . 85
5.7 Costs per set/unit based on order quantity for Bridge Board bare PCB and BOM. . 85
D.1 3D printer settings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
xii
List of Figures
2.1 Illustration of the fundamental parameters of an imaging system. . . . . . . . . . . 5
2.2 Geometry of a basic optics system using the thin-lens model. . . . . . . . . . . . . 5
2.3 Angle of view and its horizontal, vertical and diagonal components [18]. . . . . . 6
2.4 Focal length and angle of view of a “full frame” (35mm) image sensor [19]. . . . . 7
2.5 Conversion and charge transferring principles between CCD and CMOS technolo-gies [20]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.6 Comparison between rolling shutter and global shutter captures on moving target[21]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.7 Perfectly clear pattern before (left) and after (right) passing through a low resolutionlens [25]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.8 Resolving two black stripes. If the space between the stripes is too small (a) thecamera sensor will not be able to resolve them as separate stripes [25]. . . . . . . . 13
2.9 Different levels of contrast plotted as a square wave [25]. . . . . . . . . . . . . . . 14
2.10 Contrast comparison in resulting images at low (a) and high (b) frequency, and MTFplot (c) [27]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.11 SFR target with horizontal and vertical (highlighted) slanted edges. . . . . . . . . 17
2.12 Flowchart of the operations in the edge gradient analysis. . . . . . . . . . . . . . . 18
2.13 Different MTF shapes and their meaning. . . . . . . . . . . . . . . . . . . . . . . 18
2.14 Diagram of GStreamer components [38]. . . . . . . . . . . . . . . . . . . . . . . 21
2.15 Illustration of hyperspectral imaging. Each pixel in the image forms the spectrumtrough imaged wavebands. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
xiii
2.16 Comparison of band classifications in multispectral (few wide bands) and hyper-spectral (several narrow bands) models. Units are in micrometers for the multispec-tral ranges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.17 Reflectance spectrum of water, soil and vegetation. . . . . . . . . . . . . . . . . . 22
2.18 Reflectance spectrum of water, soil and vegetation together with the Landsat TMbands in the 0-2.5 µm range. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.19 (a) Tetracam Mini MCA6. (b) Sentek GEMS. (c) Sentera QUAD. (d) Parrot Se-quoia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.20 Comparison of spectral bands used in each camera. . . . . . . . . . . . . . . . . . 26
3.1 Block diagram of the UAV system board hardware. . . . . . . . . . . . . . . . . . 35
3.2 UAV system board. 90x52mm 4-layers PCB with a weight of 29 grams. A detailedmanual and user guide for the board can be found in Appendix A. . . . . . . . . . 37
3.3 Redundant Power supply module. 41x46mm 4-layers PCB with a weight of 14grams. There are four voltage regulators, from bottom left and clockwise: autopilotmain, payload, servo backup, and autopilot backup. . . . . . . . . . . . . . . . . . 39
3.4 Redundant Power supply module block diagram. Greyed blocks and dashed linesare inactive. (a) Normal power operation mode: all systems are on and poweredfrom main battery. (b) Backup power operation mode: autopilot and servos are pow-ered from backup battery; motor, payload and Electronic Speed Controller (ESC) /Battery Eliminator Circuit (BEC) are off. . . . . . . . . . . . . . . . . . . . . . . 40
3.5 Main and backup power supplies connection diagram. Both power supplies areconnected in parallel and there is a voltage difference between the two, where VMAIN
is slightly higher than VBACKUP when in normal operation mode. . . . . . . . . . . 41
3.6 Variation of the current sank by the backup voltage regulator circuit (iS) with thevoltage difference ∆V = VMAIN −VBACKUP, including the point of current inversionfor a low ∆V in which iS becomes negative. . . . . . . . . . . . . . . . . . . . . . 41
3.7 Variation of the current sank by the backup voltage regulator circuit (iS) with thevoltage difference ∆V when both voltage regulators are on (∆V =VMAIN −VBACKUP). 42
3.8 Variation of the current sank by the backup voltage regulator (iS) circuit with theoutput voltage level (VOUT ) when the backup voltage regulator is off (output in highimpedance). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.9 Paparazzi-UAV Ground Control Station (GCS) interface during a flight. . . . . . . 43
3.10 Hardkernel ODROID-C2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
xiv
3.11 UAV avionics module fully assembled with all its components. . . . . . . . . . . . 45
3.12 HobbyKing EPP FPV airframe. . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.13 Inside of the HobbyKing EPP FPV fuselage after splitting. Inside cuts are markedand nose was removed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.14 Side of the fuselage with cuts and some plastic parts glued already (right). . . . . . 47
3.15 (a) Original EPP FPV fuselage. (b) Modified fuselage (Falcon I design). . . . . . . 48
3.16 (a) Top fuselage module. (b) Assembled wing centre module. (c) Assembly attach-ment glued to the wings. (d) Wings assembly. . . . . . . . . . . . . . . . . . . . . 49
3.17 (a) Original EPP FPV tail. (b) Modified tail (Falcon I design). . . . . . . . . . . . 50
3.18 Internal wiring of the fuselage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.19 (a) Default foam nose. (b) Custom plastic nose for installing a USB camera board. 51
3.20 Rendering cut of the fuselage with internal components. . . . . . . . . . . . . . . 51
3.21 3D model of the Falcon I full assembly. . . . . . . . . . . . . . . . . . . . . . . . 52
3.22 Falcon I full assembly on the ground (a) and during flight (b). . . . . . . . . . . . 52
4.1 USB camera module ELP-USB500W02M from ELP Ailipu Technology Co. . . . 54
4.2 MTF plot for lens 5 (6mm focal length). . . . . . . . . . . . . . . . . . . . . . . . 58
4.3 Simple block diagrams of a single camera (a), and the 3-band multispectral imagingsystem (b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.4 Quantum Efficiency of the Aptina / ON Semiconductor AR0130CS image sensor. 61
4.5 Spectral response or transmissivity of the green, red, and NIR band-pass filterschosen for the system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.6 Comparison of chosen bands with bands used in other commercial multispectralcameras. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.7 From left to right; green, red, and NIR filters mounted on lens cap adapters. . . . . 63
4.8 Back of a lens with IR cut filter (left), and after IR cut filter removal (right). . . . 63
4.9 Diagram of the Gstreamer pipeline application for three ELP-USB130W01MT-B/Wcameras. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.10 3D model of custom plane nose with 3 cameras. (a) Isometric view. (b) Side cross-section view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.11 (a) Custom plane nose with 3 cameras and filters. (b) Nose installed in the plane. . 65
xv
4.12 Map of Bremner Field indicating the delimited area for flight operations. . . . . . 66
4.13 Correlation plot of alignment algorithm indicating maxima. . . . . . . . . . . . . 68
4.14 Example of pixel offset and cropping of images for alignment. . . . . . . . . . . . 69
4.15 Superposition of red, green, and NIR band images before and after alignment. . . . 69
4.16 Result of image alignment algorithm. (a) Superposition before alignment. (b) Su-perposition after alignment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.17 Example NDVI test of leaves with different levels of stress. (a) RGB image. (b)False colored NDVI image. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.18 Sample pictures of a flights at Bremner Field showing the NIR and Red band imagesand the NDVI resulting image. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.1 DMM 42BUC03-ML camera module from The Imaging Source, including S-Mountlens holder and lens. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.2 Quantum efficiency of the ON Semiconductor AR0134CS image sensor. . . . . . 75
5.3 Timing diagram of the trigger, exposure, and image readout processes. . . . . . . . 75
5.4 Camera board with S-Mount (M12) lens and optical filter installed using the 3Dprinted adapter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.5 Diagram of the basic connections between the camera array and the bridge board.The bridge board routes all USB communications into selectable upstream ports andsends trigger signals to the cameras. . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.6 Altium Designer layout view of the bridge board with top layer (red), bottom layer(blue), and silkscreen (yellow). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.7 CAD 3D model of the bridge board. . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.8 Diagram of the configurable USB HUB network. SW 1, SW 2, and SW 3 are 2-to-1bidirectional USB 2.0 switches with inputs controlled by user selectable jumpers.USB 1, USB 2, USB 3, are the isolated upstream ports, and USB M is the masterupstream USB port. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.9 Diagram of the trigger system based on a microcontroller unit (MCU). Triggersignals are generated via GPIOs. A UART connection via USB is provided forinteractively interfacing with the MCU via serial commands. . . . . . . . . . . . . 80
5.10 Multispectral imaging system hardware with 12 cameras installed. . . . . . . . . . 84
5.11 Example of a sample gray scale image (left) after applying the positive Laplacianoperator (right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
xvi
5.12 Graph of the focus measure calculated using the above algorithm as the focus isadjusted from out of focus (too far) to out of focus (too close), passing through thein-focus point. Any point outside the in-focus point is out of focus. The graphhighlights two random out of focus points for reference, one too far and one too close. 88
5.13 Sample images of the three highlighted points in the graph of Figure 5.12 . (a) Outof focus (too far). (b) In-focus. (c) Out of focus (too close). . . . . . . . . . . . . . 88
5.14 Result of aligning two images. (a) Template or reference image. (b) Test imagealigned to template image. It is possible to see the black borders filling the emptyspaces after alignment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.15 Result of image alignment algorithm. (a) Superposition before alignment. (b) Su-perposition after alignment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
xvii
List of Abbreviations
AGL Above Ground Level
BEC Battery Eliminator Circuit
CCD Charge-Coupled Devices
CMOS Complementary Metal–Oxide–Semiconductor
CPI Camera Parallel Interface
CSI Camera Serial Interface
DOF Degrees of Freedom
DOF Depth of Field
DOM Degrees of Measurement
DSC Digital Still Camera
EPP Expanded Polypropylene
ESC Electronic Speed Controller
FOV Field of View
FPS Frames Per Second
GNDVI Green Normalized Difference Vegetation Index
GPS Global Positioning System
GSD Ground Sampling Distance
GUI Graphical User Interface
IMU Inertial Measurement Unit
xviii
MCU Microcontroller Unit
MTF Modulation Transfer Function
NDRE Normalized Difference Red Edge Index
NDVI Normalized Difference Vegetation Index
NIR Near Infrared
PCB Printed Circuit Board
PLA Polylactic Acid
PMAG Primary Magnification
ROI Region of Interest
SBC Single Board Computer
SFR Spatial Frequency Response
SNR Signal-to-Noise Ratio
UAARG University of Alberta Aerial Robotics Group
UAV Unmanned Aerial Vehicle
USB Universal Serial Bus
UVC USB Video Class
V4L2 Video for Linux Two
WD Working Distance
1
Chapter 1
Introduction
Aerial and satellite photography are remote sensing techniques that have contributed to numerousdevelopments due to the extensive coverage and large amount of data that can be collected in ashort period of time. Remote sensing is based on collecting data about an object or phenomenonfrom a distance, without making physical contact with the object. Multispectral imaging is a majorremote sensing method in the field of spectroscopy that studies the amount of light that is emittedby or reflected from materials and its energy fluctuations with wavelengths. In the context of aerialand satellite multispectral imaging, data is collected by detecting the energy that is reflected fromthe Earth’s surface. Multispectral imaging focuses on a small number of carefully selected bands,usually 3 to 10 bands, that are not necessarily adjacent and tend to be 30-40 nm wide. Thesebands are usually selected for a specific application by knowing the type of materials that will bestudied and their spectral response. Aerial and satellite multispectral imaging has a wide range ofapplications, that go from research to commercial and industrial.
The Landsat Thematic Mapper (TM) [1] is a satellite mounted whisk broom multispectral scanningsensor from NASA, that has been used for extensive characterization and identification in multipleapplications [2, 3, 4, 5]. The TM has a total of 7 spectral bands, 3 in the visible spectrum and 4 inthe infrared spectrum. Each band has been carefully selected for the reflectance properties of certainmaterials of interest, such as vegetation, soil, and water. Manned aircrafts are also used for flyingmultispectral imaging equipment [6, 7]. While imaging payloads in satellites and manned aircraftshave been useful and effective, the costs and complexity associated with these implementations makethem unfeasible for small scale and low cost applications. With the increasing use of UnmannedAerial Vehicles (UAV) in recent years, and the advances and miniaturization in imaging systemstechnologies, costs and complexity have been considerably reduced.
UAVs have been successfully used for flying multispectral cameras for a variety of applications.Experiments have shown that UAV-based spectral data has a high correlation with equivalent ground
2
level measurements. Nebiker et al. [8] validated the use of a UAV mounted high-end multispectralsensor, showing high correlation with respect to ground level reference measurements, together witha low-cost commercial near-infrared (NIR) camera that showed reasonable performance.
A major application of UAV-based multispectral cameras is crop inspection and precision agricul-ture. Part of Nebiker et al. [8] experiments included plant vitality assessment and plant diseasedetection. Roosjen et al. [9] used a UAV mounted spectrometer to gather spectral anisotropy datafrom several agricultural crops as a way to measure crop growth. Candiago et al. [10] also provedthe use of UAV-based multispectral imaging for vegetation mapping and crop assessment.
Multispectral images taken from UAVs have also been successfully used for surface and terrain stud-ies, detecting soil and vegetation marks, and surface modeling. Themistocleous et al. [11] combinedUAV and ground spectral data to study the formation of crop and soil marks for archaeological ap-plications, while Turner et al. [12] used a multispectral camera together with a visible camera and athermal infrared camera to study the physiological state of Antarctic moss ecosystems. Multispectraltests from a UAV were reported by Hassan-Esfahani et at. [13] in successful estimation of surfacesoil moisture over a large field irrigated by a center pivot sprinkler system.
UAV-based multispectral imaging is a growing field that continues to have a major impact in vege-tation assessment, crop inspection, and precision agriculture. With recent advances in multispectralsensing technologies, small and lightweight commercial solutions have found their way into themarket of UAV remote sensing tools. Nonetheless, these tools are still a relatively expensive so-lution that, in some cases, lacks certain flexibility for further development. Popular multispectralcameras that are UAV compatible like the Sentek Systems GEMS Multispectral Sensor [14], theSentera QUAD [15], and the Parrot Sequoia [16] offer a small, lightweight, and highly integratedsolution. Conversely, fixed optical and functioning properties affect negatively the flexibility andpotential of these solutions.
This work aims to introduce the design and development, as well as demonstrate the use and perfor-mance of two main systems. First, a comprehensive UAV system board with redundant power supplythat integrates most UAV functions and components. Second, a custom made multispectral imagingsystem based on USB camera modules. The imaging system is a flexible hardware and softwareplatform that can be used to develop custom UAV compatible multispectral imaging solutions.
1.1 Objective of the Thesis
This thesis attempts to demonstrate the feasibility of a small scale, lightweight, low cost UAV plat-form for multispectral imaging applications. The work presented here contributes a practical designand implementation methodology for developing UAV systems, focused on an integrated UAV sys-tem board and a multi-camera imaging systems for aerial remote sensing. At first, a basic conceptual
3
3-band multispectral imaging system is developed. The presented system introduces the use of mul-tiple USB 2.0 camera modules and different optical filters for capturing simultaneous images indifferent spectral bands. Commercial solutions of UAV compatible multispectral cameras are an-alyzed and compared to the proposed system. To test and validate the system a fixed-wing UAVwas designed and built, based on a semi-custom airframe, UAV system board hardware with a novelredundant power supply system, and a single board computer (SBC) for payload. The UAV is usedto evaluate the effectiveness and performance of the system in real field multispectral imaging ap-plications. A second and more complex revision of the multispectral imaging system is developedwith a higher number of cameras and more advanced features. This second evolution serves as adevelopment platform for UAV compatible multispectral imaging applications.
1.2 Organization of the Thesis
Following this introduction is a literature review in Chapter 2, where a basic overview of imag-ing systems, remote sensing, and UAV-based multispectral applications is presented, together withprevious work pertaining to this thesis. Chapter 3 covers the development of the UAV platform,from the avionics to the airframe, including the development of a UAV system board with redun-dant power supply. Chapters 4 and 5 elaborate on the design, development, and testing of the basic3-band multispectral imaging system and the more advanced 12-band multispectral imaging systemrespectively. Finally, Chapter 6 summarizes the conclusions of this works for future developmentand application.
4
Chapter 2
Literature Review
This chapter presents a review of literature previously published relevant to the major topics coveredin this work including: imaging systems, image sensors, USB camera modules, spectral imaging,remote sensing, and the integration of unmanned aerial vehicles (UAVs) into aerial imaging remotesensing solutions with applications in multiple areas. Previous system solutions and their achieve-ments and limitations are discussed.
2.1 Imaging Systems and Cameras
An imaging system is an arrangement or setup of component that allow capturing images and han-dling them. Part of an imaging system can be to optically modify the capture, alter, store, process,and/or transmit the images. The most essential stage of any imaging system is the capturing device,which is the camera. Cameras are optical devices that capture images, either as individual still pho-tographs or as a sequence of images constituting videos. Technically speaking, cameras are remotesensing instruments as they sense subjects without any contact.
A camera consists of a combination of optical, mechanical, and in the case of digital cameras,electronic components. Traditional cameras worked based on capturing light onto a photographicplate or photographic film. Digital cameras use an electronic image sensor to capture images, andadditional electronics to transfer those images to some storage media. The present work is focusedon the use of digital cameras.
2.1.1 Fundamentals of Optics and Imaging Systems
The design and understanding of an imaging system involves a lot of optical properties, dimensionsand calculations, but there are some basic fundamental parameters and concepts (Figure 2.1) that
5
must be understood prior to any further study.
Figure 2.1: Illustration of the fundamental parameters of an imaging system.
The two main components of any camera are the image sensor and the lens. All calculations relatedto optics are based on the dimensions and characteristics of these two elements. The thin-lens model[17] is used here for the purpose of covering basic optics and imaging concepts. The thin-lensmodel simplifies ray tracing calculations by ignoring the optical effects due to the thickness of thelens. While it is an approximation, this model (Figure 2.1) illustrates lens and optics principles andintroduces the fundamental terms with which to discuss imaging systems.
Figure 2.2: Geometry of a basic optics system using the thin-lens model.
6
Sensor Size The dimensions of the camera sensor’s active area. This parameter is typically spec-ified as horizontal and vertical physical dimensions. The sensor size can be calculated by knowingnumber of horizontal and vertical pixels in the sensor and the size of a pixel. Sensor size is animportant parameter in determining the characteristics of a lens for a desired field of view.
Focal Length ( f ) The distance from the lens’s optical center to the image sensor plane when thelens is focused at infinity (Figure 2.2). Focal Length, usually stated in millimeters, is the principalmetric of a lens, and gives a measure of how strongly light converges in the optics system, assuminga positive (convex) lens.
Aperture The size of the opening through which light enters the camera. Together with focallength, it defines the cone angle of light that come to a focus in the sensor plane.
Working Distance (WD) The distance from the front of the lens to the object. For a fixed FocalLength, changing the Working Distance proportionally changes the Field of View.
Field of View (FOV) The total viewable area of the object that is imaged by the lens onto theimage sensor. This is the portion of a given scene that is captured by the camera’s sensor. It canbe expressed in terms or total area or by its horizontal, vertical, or diagonal components. FOV issometimes expressed as the angular size of the view cone, and it is called angle of view. The angleof view can also be measured horizontally, vertically, or diagonally (Figure 2.3).
Figure 2.3: Angle of view and its horizontal, vertical and diagonal components [18].
The horizontal and vertical components of the angle of view can be calculated in degrees by:
θH,V = 2arctan(
sH,V
2 f
)180π
(2.1)
7
where sH,V is the horizontal or vertical size of the sensor and f is the Focal Length. Similarly, therelationship between Focal Length, Sensor Size, Working Distance, and Field of View is defined by:
FOVH,V =sH,VWD
f(2.2)
It can be seen from (2.2) that a greater Focal Length leads to a smaller Field of View as the anglebecomes smaller. Figure 2.4 illustrates examples of different Focal Lengths, and how the angle ofview and Field of View change accordingly.
Figure 2.4: Focal length and angle of view of a “full frame” (35mm) image sensor [19].
Resolution The smallest feature size of the object that can be visibly resolved in the image. Res-olution quantifies how close lines can be to each other and still be distinguished in the final image.This is a simplified definition for a concept that is very important and widely misunderstood. Abroader and more detailed explanation of resolution will be covered later on. Concepts like sensorresolution, Modulation Transfer Function (MTF), and effective resolution of an imaging system willbe introduced.
Depth of Field (DOF) The range between the maximum and minimum working distances where itis possible to maintain acceptable focus. Depth of Field is also the total amount of object movementin the optical axis allowable while maintaining focus.
Primary Magnification (PMAG) The Primary Magnification is defined as the ratio between theSensor Size and the Field of View:
8
PMAG =sH,V
FOVH,V(2.3)
2.1.2 Image Sensor
As previously mentioned, the image sensor is the principal element of a camera, in charge of captur-ing light and transferring the information that constitutes an image. Image sensors have a sensitivecell array that converts light waves into electrical signals that are measured and interpreted in orderto produce a matrix of values that form the image.
2.1.2.1 CCD vs. CMOS Technologies
Most modern image sensors use one of two types of technologies: semiconductor charge-coupleddevices (CCD) or active pixel sensors in complementary metal–oxide–semiconductor (CMOS).
Regardless of the technology used, image sensors must first convert light into electrons that areaccumulated as charge, and then read the value of each cell. In CCD sensors, the charge is transferredfrom one pixel to another across the chip until it ends up at the periphery of the array, where it isconverted into voltage. Then, an analog-to-digital (ADC) converter turns each pixel’s voltage intoa digital value. In CMOS image sensors, there are several transistors integrated into each pixel thatamplify and transfer the charge using traditional wires. CMOS sensors are more flexible becausepixels can be read individually. Figure 2.5 shows a visual comparison of the working principlebetween CCD and CMOS sensors.
Figure 2.5: Conversion and charge transferring principles between CCD and CMOS technologies [20].
There is a number of features where CCD and CMOS image sensor are very different. CCD sensorscreate a higher-quality, low-noise image, whereas CMOS sensors are more susceptible to noise.Due to each pixel on a CMOS sensor having several transistors next to it, many of the photonswill inevitably hit these transistors and the overall light sensitivity will be lower. CCD sensorsuse a special manufacturing process that guarantees the transportation of charge across the chip
9
without distortion, leading to sensors with high fidelity. The downside is that this process consumesconsiderably more power. CCD sensors can consume up to 100 times more power that an equivalentCMOS sensor. Contrarily, CMOS chips use traditional manufacturing processes, so they can befabricated on just about any standard silicon production line. Therefore, CMOS sensors tend to beinexpensive compared to CCD sensors. Table 2.1 summarizes the comparison of some of the mostimportant features between CCD and CMOS image sensors.
Table 2.1: Features comparison between CCD and CMOS image sensors.
Feature CCD CMOS
System Noise Low ModerateResponsivity Moderate Slightly BetterPower Consumption High LowSensitivity (*) High ModerateResolution High HighCost High Low
(*) Visible and infrared bands up to approximately 1100 nm.
2.1.2.2 Rolling vs. Global Shutter
The shutter is the mechanism for exposing the sensor to light. There are two types of shutter, rollingand global, where the difference is how each controls the scanning of the image. In a rolling shuttermechanism, the image is scanned sequentially, where the pixels of the sensor are exposed to lightline by line. It usually starts from top to bottom and is more commonly found in CMOS sensors.While each pixel line collects light during the same period of time, the light collection start and endtimes for each line differ.
In a global shutter mechanism, on the other hand, all the pixels are exposed at the exact same timeand during the same time. Since light collection happens simultaneously for all pixels, the sceneis “frozen” in time, provided that the time window for light collection is short enough and that nomotion occurs during this window. If this is the case, there will be no motion blur in the resultingimage. Motion blur in likely to be present in images captured with a rolling shutter if there is anymovement occurring between the start of the first line and the end of the last line. Figure 2.6 shows acomparison of a moving object being captured by a rolling shutter sensor and a global shutter sensor.
2.1.2.3 Performance Metrics of Image Sensors
Essentially, the function of an image sensor is the conversion of light to a digital image. There aresome common metrics used for comparing how well image sensors can perform their task. This
10
Figure 2.6: Comparison between rolling shutter and global shutter captures on moving target [21].
section elaborates on some of these metrics and presents them in three categories: metrics related tothe pixel layout, metrics related to the pixel physics, and metrics related to the pixel readout [22].
Metrics Related to Pixel Layout The most common metric used for comparing image sensors isthe pixel count, usually expressed in megapixels. The number of pixels in a sensor depends on theoverall size of the chip and the area the designer has available for the pixels versus the area used forthe rest of the circuitry. When it comes to pixel count, the higher the better.
Another metric that also relates to pixel count is the pixel pitch, usually expressed in micrometers.This metric defines how much area each pixel occupies in the array. Generally, a smaller pitch meansmore pixels can be fitted. However, smaller pixels will not collect as much light, and thus might notbe suitable for fast exposure captures or low illumination applications.
Finally, one more highly used metric is the fill factor, which deals with how pixels are physicallylaid out. The fill factor is the percentage of the pixel area that is occupied by the actual photodiode. A100% fill factor would be ideal as all the pixel area would be used for light collection. Most sensorsdon’t have a 100% fill factor, and designers need to decide how much area is used for photodiodes.Nonetheless, some chips can achieve 100% fill factor by using a backside illuminated design, wherereadout circuitry is placed below the photodiode, allowing it to occupy all the pixel area [23, 24].
Metrics Related to Pixel Physics Another set of metrics by which image sensors are compared isrelated to the physics of how photons are converted into charges. When photons hit the silicon withenough energy an eletron-hole pair is formed. But this does not happen for every single photon andthe effect also differs depending on the wavelength. Quantum efficiency is a percentage measureof how many photons are converted into electron-hole pairs.
Another important metric is the dynamic range, given in decibels, which measures how accurate asensor can capture a signal under low light conditions all the way up until full capacity. A higher
11
dynamic range is desired as it will allow the sensor to operate in a wider range of light conditions.
Metrics Related to Pixel Readout A third set of metrics that is useful for image sensor com-parison has to do with the features and performance of the readout process and circuitry. One ofthe most important metrics here is the Signal-to-Noise (SNR) ratio, measured in decibels. This isa ratio of how much of the original signal is captured and present in the output, versus how muchnoise is present. The higher the SNR means there is more of the original signal present in the outputcompared to the noise level. SNR varies with light intensity, as it is typically lower for dim light andhigher for bright light.
Other important metrics are the percentage of linearity, which is how linear the pixel output is withrespect to the captured photovoltage, and the frame rate, which is a measure of how many timesper second the full pixel array can be exposed and read. Power consumption is also important. Thechoice of power supply, operating frequency and capacitance affect power consumption. Typically,a higher frequency allows a higher frame rate at the cost of higher power consumption.
Finally, the output bit-depth is another important feature. The photodiode signal that is later con-verted into voltage and read, will eventually be sampled by an analog-to-digital converter to output adigital value. The bit-depth defines the number of bits used for representing this number. The higherthe number of bit, the higher the value resolution and the more distinct values can be represented.
2.1.3 Lenses
Lenses are optical transmission devices used for focusing or dispersing light by means of refraction.Lenses are made out of transparent glass or plastic, usually with a circular shape. The two sides ofthe lens are polished surfaces, either or both of which have a concave or convex curve.
A lens can achieve its focusing effect thanks to light changing direction as it passes through thelens, crossing the interfaces between air and the lens material. This effect is called refraction, and itoccurs both where light enters the lens and where it emerges from the lens into the air.
Lenses are important because the have the valuable property of forming images of objects situatedin front of it, and are an indispensable component of any camera or imaging system. Lenses alsohave drawbacks such as non-linear distortion (which can be measured and corrected with imageprocessing) and limits to optical resolution.
2.1.4 Modulation Transfer Function
The Modulation Transfer Function (MTF) is an important and common metric for evaluating theperformance of imaging systems. MTF measures the ability of the imaging system to transfer con-
12
trast at a particular resolution from the object to the image. In order to understand the importanceand significance of the MTF, the concepts of resolution and contrast must be defined.
2.1.4.1 Resolution
A simplified definition of resolution was given in a previous subsection. A more detailed and thor-ough explanation of this concept is presented here. Resolution is a measurement of how wellan imaging system can reproduce object detail. It is often expressed in line-pairs per millimeter(lp/mm), which is a unit of frequency that originates from the testing method to measure resolution.A line-pair is a sequence of two lines with opposing contrast, that is one black and one white. Abasic target with a sequence of alternating, equally spaced black and white stripes is usually usedfor performing MTF testing. This target presents line-pairs with increasing frequency in order totest the imaging system resolving capabilities. When an imaging system captures this pattern, withperfectly clear edges, lines will have a degree of blurriness in the final image due to imaging systemlimitations (Figure 2.7). Increasing the frequency will also increase the blur. High resolution im-ages have a high level of detail and minimal blur whereas low resolution images lack fine detail andexhibit higher blur.
Figure 2.7: Perfectly clear pattern before (left) and after (right) passing through a low resolution lens [25].
As the frequency in the target increases the line-pairs become thinner and less number of pixelsare required for imaging each line-pair. If the stripes are imaged onto neighboring pixels, thenthey will appear to be one single long stripe in the image (Figure 2.8a) rather than two separatestripes (Figure 2.8b). Therefore, there is a minimum line-pair width for which the image sensor iscapable or resolving the black and white stripes. At least the distance equivalent to one pixel must bebetween two black stripes. Following this reasoning, a minimum of 2 pixels is required for imaginga line-pair, one pixel for the black stripe and one for the white stripe.
13
Figure 2.8: Resolving two black stripes. If the space between the stripes is too small (a) the camera sensorwill not be able to resolve them as separate stripes [25].
Correspondingly, this limitation is defined by the size of the pixel in the image sensor. The sensorresolution can be calculated with (2.4).
Sensor Resolution(l p/mm) =1000
2(pixel size(um))(2.4)
This is also known as the Nyquist limit. It can be seen that as pixel sizes get smaller the correspond-ing sensor resolution or Nyquist limit in lp/mm increases proportionally. Additionally, the objectresolution is calculated using the sensor resolution and the primary magnification (PMAG) of thelens (equation 2.5). It is important to note that the equation for object resolution assumes the lenshas no resolution loss. Object resolution can also be converted to um using (2.6).
Ob ject Resolution(l p/mm) = PMAG × Sensor Resolution(l p/mm) (2.5)
Ob ject Resolution(um) =1000
2(Ob ject Resolution(l p/mm))(2.6)
14
2.1.4.2 Contrast
Contrast can be understood as the relation between low and high luminance in a scene or image. Ina black and white bar stripe test target, minimum luminance will be black and maximum luminancewill be white. Any value between these two limits will be a mix of black and white (gray) lumi-nance values. A sample plot (Figure 2.9) of these value will produce a square wave with differentamplitudes according to the luminance value.
Figure 2.9: Different levels of contrast plotted as a square wave [25].
There are different way to calculate and measure contrast depending on the target used for testing.In this particular case the Michelson Contrast [26] is used. The Michelson Contrast, also knownas visibility, peak-to-peak contrast, or modulation, measures the relation of luminance in periodicpatterns where both bright and dark features are equivalent and take up similar fractions of the totalarea (e.g. sinusoidal gratings). The Michelson Contrast is defined as:
Contrast =Imax− Imin
Imax + Imin(2.7)
where Imax and Imin are the maximum and minimum luminance values respectively. The denominatoris the luminance difference while the denominator is twice the luminance average.
Considering the effect of a lens and the differences between the resulting image and the real object,as shown in Figure 2.7, the contrast plot of the resulting image will look more like a sinusoidalwave due to black and white transitions in the target not being resolved perfectly (Figure 2.10a).Moreover, as the line-pair frequency is increased, the contrast will decrease, which is seen in theamplitude of the resulting sinusoidal wave (Figure 2.10b). This effect is always present due to thelimitations of the imaging lens. For the image to appear defined, black and white must be welldefined with a minimal amount of grayscale in the transition.
15
With the concepts of resolution and contrast properly introduced, the MTF can be seen as the wayto integrate these two metrics into a single specification. As the frequency of line pairs increases onthe test target, it becomes more difficult for a lens to effectively transfer the corresponding decreasein contrast. In other words, the MTF describes the attenuation in contrast as a function of the spatialfrequency. Figure 2.10c shows a the general shape of a MTF plot, where the modulation or contrast,expressed as a percentage, decreases as the spatial frequency (in lp/mm) increases.
Spatial Frequency Units This section briefly covers the different units that used for expressingspatial frequency. There is a total of six units commonly used:
- LW/PH = Line width per picture height - Cycles/mm = Cycles per millimeter- LP/mm = Line pairs per millimeter - Cycles/pixel = Cycles per pixel- L/mm = Lines per millimeter - LP/PH = Line pairs per picture height
Table 2.2 show the conversion operations for spatial frequency units. To convert from left columnunits to top row units use the operation in the corresponding row/column intersection.
Table 2.2: Spatial frequency unit conversion chart.
LW/PH LP/mm L/mm Cycles/mm Cycles/pixel LP/PH
LW/PH x 1 / [2 x h] / h / [2 x h] / [2 x v] / 2
LP/mm x [2 x h] x 1 x 2 x 1 x p x h
L/mm x h x 0.5 x 1 x 0.5 x [p / 2] x [h / 2]
Cycles/mm x [2 x h] x 1 x 2 x 1 x p x h
Cycles/pixel x [2 x v] / p x [2 / p] / p x 1 x v
LP/PH x 2 / h x [2 / h] / h / v x 1
Where p is pixel pitch, h is picture height, and v is number of vertical pixels.
16
Figure 2.10: Contrast comparison in resulting images at low (a) and high (b) frequency, and MTF plot (c)[27].
17
2.1.4.3 Slanted-Edge MTF Measurement
The ISO 12233 standard (Photography -- Electronic still picture imaging -- Resolution and spatialfrequency responses) [28] defines a method for measuring the Modulation Transfer Function (MTF),also known as Spatial Frequency Response (SFR), based on a slanted-edge target. The slanted-edgeMTF method was developed by Burns in [29] as a modified version of traditional edge-gradientanalysis methods that used scanning microdensitometers for digital image characterization [30].Burn’s method allows to use actual image data taken from the device, and consists of a slanted,or skewed edge target (Figure 2.11), in conjunction with data processing algorithms. This methodreduces aliasing and is particularly useful for digital still cameras (DSC). This method has beenwidely and successfully applied to MTF measurements of not only digital still cameras, but alsofilm and print scanners, and CRT displays.
Work done by Williams in [31] showed the evaluation of a software implementation of the ISOprocedure and found it to provide a robust, accurate, and precise tool for quantifying MTF from adigital capture device. Williams reported the tool to be largely insensitive to edge angle variations,capable of delivering accurate MTF estimates for peak SNRs as low as 10:1, and capable of handlingill-behaved frequency morphologies caused by FIR sharpening. Refined practices in the use of thismethod have also been reported to improve performance by reducing and identifying conditions thatcan cause measurement errors [32].
Figure 2.11: SFR target with horizontal and vertical (highlighted) slanted edges.
18
The slanted-edge MTF method consists of three basic operations: acquiring an edge profile fromthe (image) data, computing the derivative in the direction of the edge, and computing the discreteFourier transform of this derivative array (Figure 2.12). The edge profile is also referred to as theregion of interest (ROI), and it can have different sizes and orientations (horizontal or vertical edge).The result is a data set that represents the MTF. A typical graph of a balanced MTF can be seen inFigure 2.10c.
Figure 2.12: Flowchart of the operations in the edge gradient analysis.
A MTF graph can have different shapes (Figure 2.13) that indicate characteristics of the imagingsystem, environment, and conditions[33]. However, for the purposes of direct performance com-parison between cameras and imaging systems, a single value that resumes MTF results has beendefined. Burns and Williams in [34] introduced the concept of sampling efficiency, which was lateradded to the ISO 12233 standard. Sampling efficiency is a measure of limiting resolution, definedas the normalized frequency at which the MTF falls to 10%.
Figure 2.13: Different MTF shapes and their meaning.
The sampling efficiency is used to calculate the resolution efficiency and the effective resolution ofan imaging system. The process starts by determining the frequency of the first 10% MTF occur-rence for both vertical and horizontal components, clipping at 0.5 cycles/pixel. Each of these valuesis normalized against the digital limit (Nyquist criteria) of 0.5 cycles/pixel to obtain the samplingefficiency of each direction (equation 2.8).
19
SamplingE f f iciencyH,V (%) =Frequency(0.1MT F)H,V (cycles/pixel)
0.5(cycles/pixel)(2.8)
The resolution efficiency is calculated as the product of the horizontal and vertical sampling effi-ciencies (equation 2.9).
ResolutionE f f iciency(%) = SamplingE f f iciencyH (%)×SamplingE f f iciencyV (%) (2.9)
Finally, the effective resolution of the imaging system can be found multiplying the delivered reso-lution (usually the image sensor resolution) with the resolution efficiency (equation 2.10).
E f f ectiveResolution(Mpx) = ResolutionE f f iciency(%)×Delivered Resolution(Mpx)(2.10)
2.1.5 USB Camera Modules
USB camera board modules are based on an image sensor, a processor that controls and reads thesensor, and an interchangeable lens. The onboard processor has a USB interface to communicate toan external computer using the USB Video Class (UVC) standard driver. These type of cameras canbe compared to webcams in terms of components, settings, and interface. Given the relatively smallnumber of components, USB camera board modules can achieve a small form factor, making thema lightweight alternative to higher-end professional cameras.
The segment of board-level cameras that are of interest in this project carry image sensors with a¼”- ½” optical format and S-Mount (M12) lenses. These lenses are small and available with a widevariety of characteristics (focal length, resolution, etc.). Interchangeable lenses increase the overallflexibility of the camera system.
2.1.5.1 USB Video Class (UVC)
The USB Video Class (UVC) is a standard that groups certain devices into a special class ofUSB device. UVC encloses devices capable of streaming video like webcams, digital camcorders,transcoders, analog video converters and still-image cameras. The advantage of UVC is that it pro-vides native “universal” and cross-platform drivers to interface with devices without depending onmanufacturer specific drivers.
20
2.1.5.2 Video for Linux Two (V4L2) API
Video for Linux Two (V4L2) API is a kernel interface that includes a collection of device drivers andan API for analog radio and video capture on Linux systems [35]. V4L2 was developed by Bill Dirksas an improved version of its predecessor Video for Linux (V4L). V4L2 provides higher flexibility,a new structure, and the possibility of interfacing with multiple applications/processes/devices at thesame time [36].
V4L2 is a high-level driver that interfaces with the lower-level UVC driver, therefore acting as amiddle layer between applications and the UVC driver.
2.1.5.3 GStreamer
GStreamer1 [37] is an open source, cross-platform, pipeline-based multimedia framework capableof linking multiple media processing systems to complete complex workflows. GStreamer providesa lot of flexibility for constructing related series of multimedia elements arranged in parallel flowscalled pipelines. These pipelines are assembled with automatic negotiation between its elements,making the construction of the workflow easy and robust.
Pipelines in GStreamer consist of a set of connected elements that can be divided in three majorgroups: data producers (sources), processing steps (e.g. formats, filters, codecs), and data consumers(sinks). The framework is based on plugins that provide codecs and functionalities.
GSreamer is written in the C programming language, but it has written bindings for various lan-guages such as Python, Vala, C++, Perl, GNU Guile, C# and Ruby. It also provides a command linetool option.
1https://gstreamer.freedesktop.org/
21
Figure 2.14: Diagram of GStreamer components [38].
2.2 Spectral Imaging
Spectral imaging is a common technique in the field of spectroscopy that studies the level of lightthat is emitted by or reflected from materials and its energy fluctuations with wavelengths. Themeasurements obtained using this technique produce simultaneous image data in multiple bands.
Two major distinctions among this technique are hyperspectral imaging and multispectral imaging.The main difference between the two is the number of bands and how narrow the bands are. Hyper-spectral imaging consists of hundreds or thousands of narrow (10-20 nm), adjacent spectral bands.This comprehensive image data set make it possible to derive a continuous spectrum of the imagecell (Figure 2.15) that can be used to identify materials.
Figure 2.15: Illustration of hyperspectral imaging. Each pixel in the image forms the spectrum trough imagedwavebands.
22
Conversely, multispectral imaging uses a smaller number of bands, usually 3 to 10 bands, that arenot necessarily adjacent and tend to be wider (30-40 nm). These bands are usually selected for aspecific application by knowing the type of materials that will be studied and their spectral response.
Figure 2.16 shows a graphical comparison between hyperspectral and multispectral imaging exam-ples.
Figure 2.16: Comparison of band classifications in multispectral (few wide bands) and hyperspectral (severalnarrow bands) models. Units are in micrometers for the multispectral ranges.
2.2.1 Vegetation, Soil and Water Reflectance
Three of the major natural materials of interest when looking at landscapes are vegetation, soil andwater. Figure 2.17 shows the reflectance spectrum of each of these materials. It can be seen thatthe three reflectance curves are fairly close in the visible spectrum (390 - 700 nm). Therefore,differentiation of these materials must be done at longer wavelengths.
0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5Wavelength (µm)
0
10
20
30
40
50
60
70
Ref
lect
ance
(%)
Dry SoilWaterGreen Vegetation
Figure 2.17: Reflectance spectrum of water, soil and vegetation.
23
Generally, water only reflects light in the visible range of the spectrum, making very distinct fromsoil and vegetation in the infrared region. The reflection of soil increases from the visible range to theinfrared range. There are different types of soil that have slightly different reflectance characteristicsdepending on its composition and the content of water and minerals. The graph above shows theaverage curve for bare ground soil. Finally, the spectral reflectance of green vegetation is verycharacteristic. Visible light is greatly absorbed by plants, specially blue and red light as they areused during photosynthesis. The curve increase abruptly in the red-edge region and then infraredlight is greatly reflected.
2.2.2 Landsat Thematic Mapper
The Thematic Mapper (TM) [1] is an advanced, multispectral scanning sensor that is part of theNASA Landsat program. It is designed for higher image resolution, sharper spectral separation,improved geometric fidelity and greater radiometric accuracy and resolution. It features a total of7 spectral bands, 3 in the visible spectrum and 4 in the infrared spectrum. Band 6 senses thermal(heat) infrared radiation in the 10.41-12.5 µm [2]. Figure 2.18 shows the graph from Figure 2.17with the addition of the Landsat TM bands up to 2.5 µm (band 6 is not shown in the graph).
Figure 2.18: Reflectance spectrum of water, soil and vegetation together with the Landsat TM bands in the0-2.5 µm range.
TM is a whisk broom scanner which takes multi-spectral images across its ground track. Most bandshave a 30 m spatial resolution, with the exception of band 6 which has a 120 m spatial resolution.Table 2.3 shows the wavelength ranges and brief description of the Landsat TM bands.
24
Table 2.3: Landsat Thematic Mapper Bands for Vegetation Monitoring.
Band WavelengthRange (nm)
Description
TM1 450 - 520 [blue-green] - maximum chlorophyll absorption. Used formonitoring aquatic ecosystems. Noisy, susceptible to scatter.
TM2 520 - 600 [green] - green vegetation reflection peak. Slight sensitivity toplant chlorophyll concentration.
TM3 630 - 690 [red] - strong chlorophyll absorption. Useful for distinguishingbetween vegetation and soil and in monitoring vegetation health.
TM4 760 - 900 [NIR] - strong water absorption. Bright reflectance for soil andvegetation. Good for defining the water/land interface.
TM5 1550 - 1750 [short-IR] - very sensitive to moisture. Used to monitorvegetation and soil moisture.
TM7 2080 - 2350 [mid-IR] - This band is also used for vegetation moisturealthough generally band 5 is preferred for that application.
TM6 10410 - 12500 [long-IR] - thermal (heat) infrared radiation.
2.2.3 Lightweight UAV-Compatible Multispectral Imaging Systems
With the increasing number of applications using multispectral remote sensing a number of multi-spectral imaging system have been developed and introduced in the market. This sections presentsa brief overview of some popular multispectral imaging systems available in the market.
The Tetracam ADC Micro [39] is a single sensor camera with an Aptina CMOS rolling shuttersensor. The camera captures the red, green and near infrared (NIR) bands using a single widebandpass filter that covers the 520-920 nm range. The camera has fixed optical parameters of 8.43mm focal length and optical aperture of f/3.2. Tetracam also has the Mini MCA4 [40], which is acamera array system. It features four cameras, 1.2 megapixels each, covering the blue, green, red,and NIR bands. There are versions of this system with 6 and 12 cameras.
The Sentek Systems GEMS Multispectral Sensor [14] is a dual camera system. It features an RGBcamera and a NIR camera, both with 1.2 megapixels sensors and a global shutter. Both camerashave a fixed focal length of 7.7 mm. This system can be acquired as an OEM PCB module or as afinal product with enclosure.
The Sentera QUAD [15] is a compact multispectral imaging system with four 1.2 megapixels globalshutter cameras. It features one RGB camera and three monochromatic single-band cameras cover-ing red, red-edge, and NIR bands. This unit integrates all four sensors into a small package.
25
(a) (b)
(c) (d)
Figure 2.19: (a) Tetracam Mini MCA6. (b) Sentek GEMS. (c) Sentera QUAD. (d) Parrot Sequoia.
Finally, the Parrot Sequoia [16] is a very popular and small unit, somewhat similar to the SenteraQUAD. It has four 1.2 megapixels global shutter single-band sensors covering the green, red, red-edge, and NIR, in addition to a 16 megapixels rolling shutter RGB camera.
Table 2.4 presents a detailed comparison of the features of each camera. At this point, it is importantto introduce a term that is used to compare the level of detail present in an image. In remote sensing,ground sampling distance (GSD) of the ground from the air or space is the distance between thecenter of two adjacent pixels measured on the ground. This measure gives a sense of how muchdistance and area are covered by pixel in the image. This measure is better as the number is lower asit relates to higher resolution. The GSD values presented in the table below correspond to a workingdistance, or altitude above ground level (AGL), of 400 ft (~122 m).
26
Table 2.4: Comparison of common multispectral imaging systems.
Camera TetracamADCMicro
TetracamMini-MCA4
SentekGEMS
SenteraQUAD
ParrotSequoia
Dimensions (mm) 75 x 59 x 33 131 x 78 x 87 127 x 89 76 x 62 x 48 59 x 41 x 28
Weight (g) 90 600 320 170 72
Number of Bands 4 4 4 4 4
Focal Length (mm) 8.43 9.6 7.7 5.0 3.98
FOV (H° x V°) 37.6 x 28.7 38.2 x 30.9 34.6 x 26.3 51.2 x 39.6 61.9 x 48.5
Sensor size (mm) 6.55 x 4.92 6.66 x 5.32 4.8 x 3.6 4.8 x 3.6 4.8 x 3.6
Resolution (pixels) 2048 x 1536 1280 x 1024 1280 x 960 1280 x 960 1280 x 960
Pixel size (µm) 3.12 5.2 3.75 3.75 3.75
GSD (cm) @120m 4.6 6.6 5.1 9.1 11.0
Price (USD) 2,995.00 9,995.00 3,500.00 4,599.00 3,500.00
Figure 2.20 shows a comparison of the bands used in each camera. It is important to note that theTetracam Mini MCA and the Sentera QUAD offer the possibility of changing the optical filters uponrequest when ordering. Nonetheless, the comparison is done based on the default configuration ofthe cameras.
Figure 2.20: Comparison of spectral bands used in each camera.
27
2.3 UAV-Based Spectral Remote Sensing
Remote sensing is the science of collecting data about an object or phenomenon from a distance,without making physical contact with the object. In the spectral scope, remote sensing refers to thecollection of data by detecting the energy that is reflected from the Earth’s surface, with instrumentsand sensors usually mounted on large scale satellites and aircrafts.
A major area in remote sensing is the field of spectroscopy, in both hyperspectral and multispectralremote sensing, which measures the energy reflectance and absorption properties of vegetation,soils, rocks, and water bodies in the natural environment, generally under solar illumination. Asmentioned before, the “traditional” approach has been to use large scale satellites and aircrafts tocollect data. But the increasing use of UAV systems, and the miniaturization and cost reduction ofspectroscopy electronics, has opened the door to UAV-mounted spectrometers, which are currentlybeing used to quickly and easily obtain spectral data of a target area.
Spectral data obtained with light-weight and compact spectrometers mounted in UAVs can achievesignificant match and a high correlation with the data obtained from similar ground-level equipment[41, 8, 42]. While ground-level measurements can generally achieve higher sensitivity and accu-racy, UAV-based measurements have shown a lower standard deviation due to the larger field ofview [41]. Therefore, it is possible to obtain reliable and accurate spectral data from UAV-mountedspectrometers without considerable losses, making this technique a faster and lower-cost alternativeto “traditional” methods of spectral data collection. The following sub-sections cover some ma-jor applications in spectral measurements and remote sensing, with a larger focus on UAV-basedhyperspectral and multispectral imaging systems.
2.3.1 Crop Inspection and Precision Agriculture
A critical element in agriculture is the inspection of crops to determine its health and quality. Thisinformation is crucial for farmers as a way to maximize crop growth, increase yields and reducecrop damage. Precision agriculture is the use of measured characteristics of soil, and the factorsthat affect it, in farming and the management of crop fields. In this regard, the most importantparameters related with crop growth are nutrient status, weed infestation, disease affectation andwater management [43].
Healthy vegetation, and more specifically their leaves, exhibit optical characteristics such as (a)low spectral reflectance at visible light (from 400 to 700 nm) due to strong absorption of thesewavelengths by the chlorophyll (pigment found in plant leaves); (b) high spectral reflectance atnear-infrared light (from 700 to 1100 nm) due to multiple scattering at the air-cell interfaces in theleaf internal tissue [44]; and (c) low spectral reflectance in the short-wave infra-red (SWIR) banddue to strong absorption by leaf constituents such as water, proteins, and carbon components [45].
28
For remote sensing purposes, spectral images can be used to extract important information on veg-etation parameters, such as canopy density, height, and clumping [46], nitrogen content [47], leafarea index [48], foliar water content, and soil properties, such as soil surface roughness [49] and soilmoisture content [50].
In evaluating and interpreting remote sensing measurements of vegetation, several indices are usedas graphical indicators of the existence of live green vegetation. The first one is the NormalizedDifference Vegetation Index (NDVI), calculated by
NDV I =NIR−RNIR+R
(2.11)
where R and NIR are the spectral reflectance measurements acquired in the visible (red) and near-infrared regions, respectively. Other indices like the Green Normalized Difference Vegetation Index(GNDVI) and the Normalized Difference Red Edge Index (NDRE) are also used for the same pur-pose while providing different levels of information. In the GNDVI (2.12), the red part of thestandard NDVI is replaced with a very specific band of light in the green range (G).
GNDV I =NIR−GNIR+G
(2.12)
Similarly, in the NDRE (2.13), the red part of the standard NDVI is replaced with a red edge (RE)spectrum centered around 715 nm.
NDRE =NIR−RENIR+RE
(2.13)
There is a large number of additional indices that have been designed to extract specific informa-tion. It is important to note that crops spectral measurements and therefore vegetation indices infield conditions are affected by interference factors such as mixture of vegetation, soil-brightness,environmental effects, shadow, soil color and moisture [51].
Hyperspectral imaging systems have been successfully used in the past to study crops using vege-tation indices (VI). Part of Nebiker et al. [8] experiments included plant vitality assessment using ahigh-end spectrometer and a low-cost consumer-grade NIR camera mounted on a UAV. In this work,a first study found a correlation between rape yield crops and NDVI values of 0.78 for treated plotsand 0.35 for untreated plots. Moreover, vitality differences among all the crops were identified andit was found that the use of fungicides prolonged plant activity up to harvest time. A second part ofthe vitality assessment study was done using different species of barley crops. In this study, correla-tions varied considerably depending on the index used (NDVI, GNDVI and NDRE) and the barleyspecie. Some species showed a low correlation with one or even two of the three indices, whileother species showed a high correlation with all three indices. These results indicated that indicesand species under study must be correctly matched in order to get the most accurate measurement.
29
Vegetation indices extracted from multispectral and hyperspectral images can be used with a highlevel of accuracy in estimating vegetation vitality and density [10] and by performing successivemeasurements over a period of time it is possible to estimate the growth rate of crops [52, 9].
Previous works have also demonstrated it is possible to measure the level of nitrogen in crops us-ing hyperspectral imaging systems. Most nitrogen levels in plant leaves is found in chlorophyllmolecules, therefore, there is a strong relationship between leaf nitrogen content and leaf chloro-phyll content [53]. An increase in visual reflectance of plants is related to the decrease in chlorophyllcontent as a result of low nitrogen supply, whereas an increase in near-infrared (NIR) reflectance isassociated to an increase in leaf area and green biomass. Therefore, by measuring leaf reflectance itis possible to predict crop nitrogen levels and therefore estimate the level of vegetation, health andcrop growth.
In experiments done by Zheng et al. [54] and Agüera et al. [43], UAV-mounted spectrometerswere used to collect data from rice and sunflower crop fields, finding strong relationships betweenvegetation indices values and the levels of nitrogen in the crop fields. Pölönen et al. [55] werealso able to estimate biomass and nitrogen content of wheat and barley fields using a UAV-mountedhyperspectral camera. In this study, four vegetation indices were tested, with FVA (linear) givingthe best correlation of 0.814 and 0.72 for biomass and nitrogen content respectively.
It is also possible to use hyperspectral images and vegetation indices to analyze, detect and trackplant stress levels. One of the most common sources of plant stress is diseases and infestations. Atearly stages, plants react to the presence of a pathogen with some physiological mechanism such asa decrease in the photosynthesis rate, which leads to an increase of fluorescence and heat emission[44]. At a later stage, pathogens cause a reduction of chlorophyll content in leaves due to necroticor chlorotic lesions [56]. A decrease in chlorophyll increases reflectance in the visible light bandand causes a shift of the red-edge position in the spectrum. At the canopy scale, infections canchange canopy density and leaf area. While these effects do not provide unambiguous indicationof a specific stress, they can provide information for early detection and anticipation of non-normalconditions.
Nebiker et al. [8] captured spectral images of an onion crop field using a UAV on two differentdates and found a substantial difference in NDVI values that were used to identify an infestationof thrips. Similarly, an infestation of blight was also detected by analyzing the changes in NDVIvalues from two flights over a field of potato crops. Näsi et al. [57] developed a processing approachbased on UAV-based hyperspectral imagery that was used to identify spruce trees suffering from aninfestation of bark beetle, and achieved a accuracy of 90% in separation between healthy and deadtrees. Similarly, Garcia-Ruiz et. al. [58] achieved an accuracy of 85% in the detection and classifi-cation of Huanglongbing-infected citrus trees using high resolution hyperspectral images taken froma UAV. It was also found that the early stage branch-level disease could be detected by performinglow altitude flights.
30
Another commonly found form of plant stress is due to lack of water, which directly affects pho-tosynthesis and chlorophyll content in leaves [59, 60], and therefore affects the spectral reflectancesignature of plants. Zarco-Tejada et. al. [61] were able to track stress levels in a citrus crop us-ing thermal and hyperspectral imagery obtained with a UAV, identifying areas where water wasdeficient.
The successful deployment of UAV-mounted hyperspectral cameras and the extraction of accurateinformation for the assessment of plants vitality and growth and the detection of plant stress condi-tions make it possible to used these systems for advanced precision agriculture and remote sensingof vegetation and crops.
2.3.2 Surface and Soil Study
Hyperspectral images taken from UAVs have also been successfully used for surface and terrainstudies, detecting soil and vegetation marks, and surface modeling. Themistocleous et al. [11]combined UAV and ground hyperspectral data to study the formation of crop and soil marks forarchaeological applications, while Honkavaara et al. [62] were able to measure the 3-D surfacemodel and surface moisture of a peat production area using light-weight hyperspectral frame formatcameras from a UAV. Turner et al. [12] also used a multispectral camera together with a visiblecamera and a thermal infrared camera to study the physiological state of Antarctic moss ecosystems.
Dry soil presents a higher albedo due to the high reflectance of its constituents. Moist soil willexhibit a low albedo due to water absorbing light. Bryant et al. [63] evaluated hyperspectral mea-surements for monitoring surface soil moisture, and found a strong correlation between reflectanceand soil moisture within 3 cm of depth. Similarly, Finn et al. [64]compared soil moisture measure-ments between airborne hyperspectral data and equivalent ground readings. A strong correlation(R2 greater than 0.7) was found in the 1000-1600 nm band for a 5.08 cm (2 inches) depth. Multi-spectral tests from a UAV were reported by Hassan-Esfahani et at. [13] in successful estimation ofsurface soil moisture over a large field irrigated by a center pivot sprinkler system. The multispectraldata combined with an artificial neural network developed to quantify effectiveness of the methodreported a high correlation (R2) of 0.77.
2.3.3 Surface Hydrocarbon Detection
The use of hyperspectral and multispectral imaging for the detection of hydrocarbons in the surface,thus being able to identify such hydrocarbons, can be a considerable improvement for environmentaland industry monitoring. Just the petroleum industry is responsible for a large amount of leaks andspills that commonly occur during transportation, storage, and manipulation of, not just petroleum,but also its related products. Therefore, it is of great interest to be able to detect and prevent these
31
events. Allen and Krekeler [65] performed a series of experiments with light, intermediate and heavycrude, and also some common refined products (E85 fuel, gasoline, diesel, and motor oil) over arange of substrates that include concrete, asphalt, gravel, grass, sands, and clays. They were ableto detect and identify these hydrocarbons from each other and water using an ASD FieldSpec Pro-FR2 spectrometer. Airborne hyperspectral data has also been collected and used for the detectionof hydrocarbons. Hörig et al. [66] flew a HyMap hyperspectral sensor and detected hydrocarbonson the ground surface. It was shown that hydrocarbons are characterized by a typical absorptionmaxima at 1730 nm and 2310 nm.
Analysis of data collected using the Airborne Visible Infrared Imaging Spectrometer (AVIRIS) dur-ing the Deepwater Horizon spill provided a quantitative way of estimating oil thickness and oil-waterratio [67]. Oil thickness and emulsion ratio maps were generated by Clark et al. [68] using the oilabsorption bands, which later were used to estimate oil volume.
Work has also been done in detecting and estimating the bitumen content of oil sands using infraredspectra. Rivard et al. [69] demonstrated that hydrocarbons in Athabasca oil sands have distinctfeatures in the infrared spectra region. In another experiment, they were able to estimate the bitumencontent of oil sand ore samples using hyperspectral reflectance spectra in the 300-2500 nm range[70].
2.3.4 Environmental Monitoring
The general study of environmental phenomena and the effect of human activity in the environmentcan also be assessed using spectral imagery techniques. Hausamann et al. [71] performed studiesoriented towards the detection of gas leaks from pipeline installations using UAV spectral remotesensors. This type of monitoring is of relevant importance in order to detect problems early andprevent negative environmental impacts. Tuominen et al. [72] reported using hyperspectral remotesensing techniques for monitoring nuclear fuel repository sites in search for environmental changes.
Experiments have also been performed using UAV spectral measurements in volcanology. McGo-nigle et al. [73] used a UAV mounted spectrometer to study a volcano in Italy and were able tosuccessfully measure carbon dioxide flux in the volcanic gas.
2.3.5 Archaeology
Spectral imaging is also an important data collection method used in archaeology. Studies have beendone in ways to extract archaeological information using vegetation indices and crop marks [74, 75].Furthermore, the increasing use of UAVs for archaeological remote sensing provides a quick docu-mentation, monitoring, and visualization tool that is easy, quick, and comes at a low cost. Cowleyet al. [76] put together a comprehensive list of UAV based surveying methods for archaeological
32
studies and highlights the usefulness and beneficial impacts of this technology in the field. A studycase in this publication was done by Moriarty [77], in which a UAV based multispectral camera wasused for identifying and monitoring crop marks matched with archaeological regions of interest. Thetracking of changes in the crop marks allowed the classification of all the regions across the area ofstudy. Similarly, Themistocleous et al. [11] acquired spectral data from a UAV and were able toidentify the formation of crop and soil marks tied to archaeological remains. With this method, itwas possible to attain a greater level of accuracy in delimiting the areas of interest.
33
Chapter 3
UAV Development
This chapter presents the development of a fixed-wing UAV as part of the setup stage for testing UAVpayloads. The intention was to develop a reliable UAV that could serve as a testing platform for longterm research applications. While off-the-shelf airframes provide certain level of flexibility, in mostcases it is not enough and modifications or adaptations are usually needed. The decision to redesignan existent off-the-shelf UAV platform comes from the need to have a high level of modularity sothat it is flexible enough to be used in different applications. The goal was to study, modify, andimprove the fixed-wing UAV platform used by the University of Alberta Aerial Robotics Group(UAARG). Making this the starting point helped leveraging the existent knowledge and experience,thus reducing the development time. The airframe is complemented with a customized UAV systemboard that integrates all UAV essential navigation and control functions, together with a redundantpower supply board that provides main and backup power regulation.
The following sections describe the design and development of the airframe of a fixed-wing UAVand a fully integrated UAV system board with redundant power supply. The UAV platform describedin this chapter, both airframe and electronics, is used for testing the payload systems developed inthe subsequent chapters, and is part of the complete UAV solution presented in this thesis.
3.1 UAV Electronic System
A fundamental feature of many UAVs is the autonomous navigation system, which is based onan onboard autopilot computer that controls and guides the UAV in flight without assistance fromhuman operators. Increasing efforts have been made in both hardware and software in order toproduce reliable and robust autopilot platform solutions. Currently, there are a large number ofhardware development boards available that can be used with a wide range of fixed-wing and rotary-wing airframes. Besides a hardware platform, these solutions bring a full software package for tasks
34
like programming, calibration, tuning, simulation, and monitoring. Chao et al. [78] made a surveyon existing autopilot platforms at the moment, comparing the main features of both proprietary andopen-source solutions. More recently, Lim et al. [79] and Ebeid et al. [80] made an extensivecomparison of available open-source autopilot hardware and software platforms. These completesolutions can make the task of building and flying a UAV relatively easy and fast.
Some hardware platforms are designed to be a minimal solution, relying on additional dedicatedboards or modules that are wired together, giving some flexibility to the user for assembling a cus-tomized final system [81]. Other approaches are focused on full integration of components into asingle board, aiming for simplicity and reduction of wiring [82, 83]. A minimal autopilot systemincludes an onboard processor and attitude sensors (accelerometer, gyroscope and magnetometer).Additional standard components that are usually needed (depending on the airframe and application)and connected to the autopilot base board are: motor(s), servo(s), Global Positioning System (GPS)unit, pressure sensor, RF telemetry transceiver, and remote control (RC) receiver for manual flightcontrol.
Another feature that is not commonly found in a UAVs electronic system is a redundant powersupply that allows the UAV to recover from a main power loss. This feature has been explored insome concept development designs [83] and is found in some high-end proprietary solutions [84].
This section presents a comprehensive autopilot system board for UAVs [85] that integrates all essen-tial sensors and components. The features and parts of the board are described in detail, emphasizingthe high level of integration that simplifies the assembly, reduces wiring and weight, and increasesreliability.
A secondary fully compatible, detachable, and novel board is also presented, that provides powerregulation and redundancy. The architecture of this module is presented in detail, highlighting theparallel power supply approach used, which results in minimal power loss and instant switchingfrom main to backup power supply.
3.1.1 UAV System Board
The first module is the system board which incorporates all common UAV electronics, including ba-sic sensors, Microcontroller Unit (MCU) and communication ports. Additionally, the board featuresan on-board data logger with a microSD card port, gates to enable/disable Pulse Width Modulation(PWM) control outputs (typically wired to an enable switch), and an expansion port that gives ac-cess to most MCU pins. Important features are described in more detail in the following sections. Adetailed user guide for this board can be found in Appendix A.
The UAV system board was designed to be fully compatible with the existing Lisa MX v2.1 classautopilot board designed by 1BitSquared, which is similar to the Lisa M autopilot board [86] with
35
an upgraded MCU. This boards are supported in the Paparazzi-UAV project [87, 86, 88], which isan open-source ground and airborne software architecture UAS platform. This compatibility savesany need to develop custom software and gives the possibility to reuse the existing firmware filesdesigned for the Lisa MX board. The Lisa MX was selected mainly because of its powerful MCU.Figure 3.1 shows the block diagram of the autopilot board hardware.
Figure 3.1: Block diagram of the UAV system board hardware.
3.1.1.1 Microcontroller Unit (MCU)
The UAV system board is built around the STMicroelectronics STM32F405VGT6, a 32-bit ARM®Cortex® M4 MCU running at 168MHz with 192KB RAM, 1024KB Flash, and Floating Point Unit(FPU). The MCU includes 82 I/O, 10 Timers, and CAN, I2C, I2S, SPI and UART communicationinterfaces in a LQFP-100 package.
3.1.1.2 Inertial Measurement Unit (IMU)
The 10 Degrees-of-Freedom (DOF) or 10 Degrees-of-Measurement (DOM) Inertial MeasurementUnit (IMU) is comprised by an integrated 3-axis accelerometer and gyroscope (Invensense MPU6000),a 3-axis magnetometer (Honeywell HMC5883L) and a barometer (Measurement Specialties MS5611-01BA03). The accelerometer/gyroscope and barometer are connected via SPI interface to the MCU,and the magnetometer is connected to the accelerometer/gyroscope via I2C interface. The ac-celerometer/gyroscope acts as a master with the magnetometer, pulling the data and saving it locally,allowing the MCU to extract data from these three sensors directly from the MPU6000 via SPI.
36
3.1.1.3 GPS
The system includes a uBlox MAX-7Q GPS module with a super-capacitor for hot start and cor-responding antenna matching circuitry. The GPS works in the 1575.42Mhz band and the antennacircuitry includes a SAW filter and optional Low Noise Amplifier (LNA) that can be bypassed ifused with an active antenna. The board supports antennas with either uFL or SMA connectors. Theconnectors can be selected using zero ohm jumper resistors. When power is off, the super-capacitoris capable of keeping the GPS module in backup battery mode for approximately 20 minutes toallow for hot start [81]. The GPS is connected to the MCU via a UART communication interface.Optionally, when an external GPS unit is used, the on-board GPS can be powered off and discon-nected from the MCU by means of on-board jumpers. The external GPS can then be powered andconnected to the MCU UART port using the corresponding connector.
3.1.1.4 XBee and Data-logger
A socket is included for connecting an XBee RF transceiver module for telemetry. The XBee isconnected to the MCU via UART communication interface. A secondary MCU based on an AtmelAVR 8-bits 16Mhz ATMEGA328P logs the telemetry data from the serial port to a micro-SD card.This data-logger uses a modified version of the open-source “OpenLog” firmware that was adaptedfor Paparazzi by the Paparazzi team. The data-logger starts running and saving data automaticallyafter power up. The board includes a jumper for powering on/off the data-logger.
3.1.1.5 PWM Outputs
There are eight PWM outputs that can be used to control motors and servos. Output signals runthrough gates and each signal can be enabled/disabled by connecting a switch or a jumper to oneof two user control inputs that can be accessed from an on-board connector. This functionalityprovides an easy way to have kill-switches for turning off PWM signals to the motors and servosindependently. This feature is useful for debugging, programming and tuning output signals whilekeeping the motors off. This also serves as a safety feature that can prevent human injury, damageto servos while programming, accidental motor spinning and fly-away incidents. PWM outputs canbe assigned to either the motor or the servo group by using solder jumpers (one for each output).
3.1.1.6 Power Regulation
The on-board power supply is split into two 3.3V Low Drop-Out (LDO) linear voltage regulatorswith +5V input. The first voltage regulator (Texas Instruments LP2992AIM5-3.3) powers the ac-celerometer, gyroscope, magnetometer and barometer sensors, so as to reduce noise in the measure-
37
ments, while the second voltage regulator (Texas Instruments LM3940IMPX-3.3) powers to the restof the board. The latter is capable of delivering up to 1A of power.
3.1.1.7 Additional Features
Finally, the UAV system board has additional I2C, SPI, CAN, and UART ports for external com-munications, analog inputs, GPIOs, bind button, status LEDs, and connection for two radio control(RC) receivers. Some communication ports have dedicated connectors and most MCU pins are ac-cessible through a 48-pin expansion header. The expansion header provides the opportunity for thedesign and use of daughter boards and “shields”, increasing flexibility and expanding functionalities.The features of the UAV system board are summarized in table 3.1.
Figure 3.2: UAV system board. 90x52mm 4-layers PCB with a weight of 29 grams. A detailed manual anduser guide for the board can be found in Appendix A.
38
Table 3.1: UAV System Board Features
Characteristic Specifications
Microcontroller STM32F405VGT6
10 DOF IMU 3 axis accelerometer3 axis gyroscope3 axis magnetometerbarometer
Remote Control Two UART ports for remote control receivers
GPS One uBlox MAX-7Q GPS unitThe GPS on-board circuitry includes:- SAW filter- Optional LNA- Selectable uFL or SMA connector for GPS antenna
Connectivity 8 PWM, 1 JTAG, 1 SPI, 2 I2C, 1 CAN, 3 GPIO, 3 ADC, 1UART
Data-logger Secondary MCU and micro-SD card socket for telemetrydata-logging
Power 3.3V@1A voltage regulator
User Interface Bind buttonEight status LEDs
Size and weight 90x52 mm, 29 grams, 4 layers PCB
3.1.2 Redundant Power Supply Board
The second module is a redundant power supply board comprised of main and backup voltage reg-ulators, shown in Figure 3.3. This module was designed as a secondary and optional board thatprovides main power for regular operation and backup power for emergency operation. Both mainand backup run on separate battery packs. The purpose of the backup power supply is to providepower to the UAV system and servos in case the main battery or main power supply fails, so that acontrolled emergency landing of a fixed-wing UAV can be performed. If the backup power is notrequired the board can be easily replaced by any 5V power supply with enough current to power theUAV system board.
39
Figure 3.3: Redundant Power supply module. 41x46mm 4-layers PCB with a weight of 14 grams. There arefour voltage regulators, from bottom left and clockwise: autopilot main, payload, servo backup, and autopilotbackup.
Each voltage regulator is based on the Texas Instruments LM43603PWPT synchronous step-downDC-DC buck converter, which is capable of delivering up to 3 A from an input voltage that can goup to 36 V, suitable for up to an 8 series cell lithium polymer battery. Furthermore, each voltageregulator has over-temperature thermal shutdown protection, output short circuit protection, andadjustable switching frequency. There is also a low battery voltage protection feature that shutsdown the regulator if the input voltage goes under a certain threshold, thus protecting the batteryfrom over-discharge damage. This threshold is adjustable by means of an on-board voltage divider.The output voltage is also variable and can be adjusted by replacing a single resistor. This designuses backup voltage regulators placed in parallel with the main voltage supply, taking advantageof the high impedance of the output of the LM43603PWPT when it is off. A small difference inregulator set point (e.g. 100mV) keeps the backup power supply in standby as described below.
Figure 3.4 shows the connection diagram of the main and backup power system with the rest of theairplane electronics in the two possible operation modes. Under normal power operation (Figure3.4a) only the main battery is used, all systems are on and the two backup regulators (autopilot andservo) are inactive.
40
Autopilot Main Supply
Autopilot Backup Power Supply
Autopilot
Motor
MainBattery
BackupBattery
Servos
Servo Backup Power Supply
Payload
ESC BEC Supply
(a)
Autopilot Main Supply
Autopilot Backup Power Supply
Autopilot
Motor
MainBattery
BackupBattery
Servos
Servo Backup Power Supply
Payload
ESC BEC Supply
(b)
Figure 3.4: Redundant Power supply module block diagram. Greyed blocks and dashed lines are inactive. (a)Normal power operation mode: all systems are on and powered from main battery. (b) Backup power operationmode: autopilot and servos are powered from backup battery; motor, payload and Electronic Speed Controller(ESC) / Battery Eliminator Circuit (BEC) are off.
In the event of main battery or main power supply failure, the system enters in backup power op-eration (Figure 3.4b), where the backup battery is used. System board and servos are powered bythe backup regulators. The payload, ESC and motor(s) remain off in this mode. When running onbackup power for the system board (including telemetry radio and RC radio receiver) and servos,an emergency glide landing for a fixed-wing UAV can be performed without motor power. For thecase of a main power supply failure in a rotary aircraft, telemetry, including situational awarenessand GPS location for recovery would be maintained.
3.1.2.1 Principle of Operation
The operation of the redundant power supply is based on the the use of two voltage regulatorsconnected in parallel (Figure 3.5), with the main regulator having a slightly higher voltage (VMAIN)than the backup regulator (VBACKUP), making the latter stay inactive while the former sources all thecurrent. In the case when the main regulator shuts down, its output goes to high impedance and thebackup regulator starts to source the current. VOUT will always be the higher of the two supplies’voltages.
When both voltage regulators are on (normal operation mode), meaning the backup regulator isinactive due to the main regulator having a higher output voltage, there is a current flowing from themain regulator to the backup regulator. There is also a current drawn from the backup battery. Letus define these two currents as iS, the current sunk by the backup voltage regulator while inactive,
41
BackupBattery
MainBattery
VOUT
VMAIN
VBACKUP
iS
iB
Main Power Supply
Backup Power Supply
Figure 3.5: Main and backup power supplies connection diagram. Both power supplies are connected inparallel and there is a voltage difference between the two, where VMAIN is slightly higher than VBACKUP whenin normal operation mode.
and iB, the current drawn by the inactive backup voltage regulator from the backup battery (batterydrain current). It is desired for these two currents to be as small as possible so that neither the mainnor the backup batteries capacity is affected in a considerable way.
Current iB is the result of the constant current drawn by the LM43603PWPT circuit when it isinactive (approx. 65µA) plus the current dissipated in the voltage divider, which depends on thevalue of the resistors and the battery voltage. The value of the resistors should be in the range of0.1-1MΩ, in order to keep iB under 100µA. Current iS varies depending on the voltage difference∆V = VMAIN −VBACKUP. As shown in Figure 3.6, there is a threshold in the values of ∆V underwhich iS becomes negative, meaning that the backup voltage regulator is active and starts sourcingcurrent together with the main voltage regulator. It is desired to keep ∆V above such threshold inorder to keep the backup voltage regulator inactive and iS positive.
0.04 0.06 0.08 0.1 0.12 0.14Voltage Difference ∆V (V)
-5000
-4000
-3000
-2000
-1000
0
1000
Cur
rent
i S (µ
A)
Figure 3.6: Variation of the current sank by the backup voltage regulator circuit (iS) with the voltage difference∆V =VMAIN −VBACKUP, including the point of current inversion for a low ∆V in which iS becomes negative.
42
Therefore, for any voltage difference VMAIN −VBACKUP > 70mV the backup voltage regulator willremain inactive and the main voltage regulator will source all the current. Figure 3.7 shows that fora wide range of voltage differences ∆V , from 75mV to 130mV, the value of iS varies slightly with alinear behavior from 72.6µA to 73.2µA.
0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.14Voltage Difference ∆V (V)
72.6
72.7
72.8
72.9
73
73.1
73.2
Cur
rent
i S (µ
A)
Figure 3.7: Variation of the current sank by the backup voltage regulator circuit (iS) with the voltage difference∆V when both voltage regulators are on (∆V =VMAIN −VBACKUP).
In the case of the backup voltage regulator being off (battery depleted or disconnected, or regulatorforced off), its output goes to high impedance and the current iS drops while still varying dependingon the voltage seen at the output. Note that this scenario is the same as when the main battery isdepleted and the main voltage regulator shuts down, enabling the backup voltage regulator (in thiscase iS would be negative according to the direction defined in Figure 3.5). Figure 3.8 shows therelation between iS and the output voltage VOUT .
5.06 5.08 5.1 5.12 5.14 5.16 5.18 5.2 5.22Output Voltage VOUT (V)
55.4
55.6
55.8
56
56.2
56.4
56.6
56.8
Cur
rent
i S (µ
A)
Figure 3.8: Variation of the current sank by the backup voltage regulator (iS) circuit with the output voltagelevel (VOUT ) when the backup voltage regulator is off (output in high impedance).
43
3.1.3 Firmware and Ground Control Station (GCS)
As mentioned before, the UAV system board is compatible with the Paparazzi-UAV platform, andthe same drivers, configuration files and code used for the Lisa MX can be used with this board.Flashing of the MCU can be done using any ARM Cortex MCU programmer, although the BlackMagic Probe is supported by the software. The Black Magic Probe is a JTAG and SWD adapterdesigned by 1BitSquared and Black Sphere Technologies for programming and debugging ARMCortex MCUs.
The Paparazzi-UAV software platform offers Ground Control Station (GCS), airborne code, flightplans, mission control, and communications [88] segments integrated in an open-source fully cus-tomizable distribution. The GCS (Figure 3.9) is an essential development, debugging, and visualiza-tion tool that allows interfacing easily with the UAV, both on the ground and during flight. It allowscontinuous and real time monitoring and logging of all variables including raw sensor readings,remote commands, and control outputs.
Figure 3.9: Paparazzi-UAV Ground Control Station (GCS) interface during a flight.
44
3.2 Payload Computer
The UAV is equipped with a single board computer (SBC) used for interfacing with payload andprocess real-time payload data. The ODROID-C2 was chosen for this task. The ODROID-C2 fromSouth Korean manufacturer Hardkernel is a small and lightweight 64-bit quad-core SBC.
Figure 3.10: Hardkernel ODROID-C2.
The ODROID-C2 was loaded with Ubuntu 16.04. USB ports were used for interfacing with USBcamera modules. A UART port accessible from the side male header was connected to the telemetryoutput of the system board for internal logging of data. The ODROID-C2 provides access to theterminal console via a UART link. This link was used to control the ODROID via command linefrom an external computer. Table 3.2 summarizes the technical specifications of the ODROID-C2.
Table 3.2: Technical Specifications of the ODROID-C2
- Amlogic ARM® Cortex®-A53(ARMv8) 1.5Ghz quad core CPUs
- Mali™-450 GPU (3 Pixel-processors + 2 Vertex shader processors)
- 2 Gbyte DDR3 SDRAM
- Gigabit Ethernet
- HDMI 2.0 4K/60Hz display
- H.265 4K/60FPS and H.264 4K/30FPS capable VPU
- 40pin GPIOs + 7pin I2S
- eMMC5.0 HS400 Flash Storage slot / UHS-1 SDR50 MicroSD Card slot
- USB 2.0 Host x 4, USB OTG x 1 (power + data capable)
- Infrared(IR) Receiver
- Ubuntu 16.04 or Android 6.0 Marshmallow based on Kernel 3.14LTS
45
3.3 UAV Avionics Module
The final avionics module for the Falcon I airframe (covered in the next section) is a single assemblythat can easily slide in and out of the plane (Figure 3.11). It contains all the main electronicsincluding the UAV system board with the redundant power supply, a control switches board usedto power on/off the boards and enable/disable the motor and servos, and the payload computer(ODROID). A minimal wiring is necessary for connections between the ODROID and the UAVsystem board, and USB cables from the ODROID are used for payload connection.
Figure 3.11: UAV avionics module fully assembled with all its components.
3.4 Airframe
The starting point for the airframe design was set based on the work done by the University ofAlberta Aerial Robotics Group (UAARG). UAARG is a student vehicle project in the Faculty ofEngineering of the University of Alberta. Their project is to design, build, test, and take to competi-tion a fully autonomous UAV.
At the time, UAARG was using the HobbyKing EPP FPV airframe for their planes (Figure 3.12).The EPP FPV is a lightweight fixed-wing airframe made out of mostly expanded polypropylene(EPP). The fully assembled airframe has a total length of 1320mm, a wingspan of 1800mm, andweighs approximately 1000 grams just the airframe (no electronics, batteries, motors, etc).
UAARG would build the airframe, and then install all the avionics (autonomous navigation and con-trol electronics) and payload systems. The first step was analyzing the system design, construction,setup and deployment procedures. This analysis, together with experience information provided bythe team members, produced a list of problems and potential improvements.
46
Figure 3.12: HobbyKing EPP FPV airframe.
One of the major issues that was noted was inherent to the airframe design. The assembly processinvolves putting together the left and right wings, then placing the wings on top of the fuselage, alignvisually, and finally secure them with rubber bands. This setup is not reliable since the alignment ofthe wings with respect to the fuselage is done visually. It is almost impossible to do the placementthe exact same way every time. While this might not represent a major problem during manualflight, where the pilot can compensate these inconsistencies by trimming the plane and adjusting toit, it does represent a considerable problem when it comes to autonomous navigation. One of thefirst steps in setting up the autonomous navigation system is to tune it according to the behaviorand response of the plane. Once it has been tuned, the autonomous navigation system will be ableto accurately control the plane. Tuning parameters are specific for each system. Therefore, if theassembly of the plane is not constant and repeatable, the tuning would have to be performed everytime, otherwise the autonomous navigation system would underperform. Tuning involves time andeffort, and it should be a one-time only step for as long as the plane is kept the same.
Another problem found in the airframe was the accessibility of the space inside the fuselage. Withonly two access points, top and front, installation and handling of the electronics is limited.
Considering that UAARG had been using the EPP FPV for 3 years at that time and that it has provento be stable and easy to control, the plan was to keep using this airframe and take advantage of theexistent knowledge and experience. With the aforementioned issues in mind a redesign of the planewas done. The aim was to solve the problems and produce a customized version of the airframe witha reliable and repeatable assembly, and optimize the space inside the fuselage for better use.
47
3.4.1 Falcon I
The Falcon I is the first iteration of the EPP FPV modified airframe. Measurements were takenfrom the original airframe and drawn in SolidWorks. Then, modifications were introduced on topof the base model. While the EPP material is very forgiving when it comes to hard landings andminor crashes, as it absorbs impacts very well, a more rigid material is needed for some parts likejoints, attachments, etc. Therefore, some parts of the plane were replaced with plastic parts so asto strengthen and reinforce key structural points. Some of the plastic parts were designed to allowextra space were to put electronics, wiring, etc.
One of the most important aspects of the redesign was to split the fuselage in two, giving access tothe inside space. This way, it was possible to work, modify, and optimize the inside of the fuselageto allocate the electronics efficiently (Figure 3.13). Some cuts were done in the inside of the fuselageto allocate more space (Figure 3.14), without affecting the overall structural strength of the body.
Figure 3.13: Inside of the HobbyKing EPP FPV fuselage after splitting. Inside cuts are marked and nose wasremoved.
Figure 3.14: Side of the fuselage with cuts and some plastic parts glued already (right).
48
Figure 3.15 shows a comparison of the unmodified EPP FPV fuselage with the redesigned version.The parts shown in blue in the redesigns are made out of plastic. All plastic parts were designedto be 3D printed using polylactic acid (PLA) as the printing material. PLA is a biodegradable andbioactive thermoplastic aliphatic polyester commonly used in 3D printers. 3D printer settings usedfor these parts can be found in Appendix D.
(a) (b)
Figure 3.15: (a) Original EPP FPV fuselage. (b) Modified fuselage (Falcon I design).
A top fuselage module (Figure 3.16) was designed to fit the wings centre module and also be able touse the space inside of it to fit the GPS/GPS antenna and external magnetometer. The wings centremodule (Figure 3.16) is a middle structure that goes between the wings and holds them togetherusing rods for alignment and an attachment piece that is glued to the wings. It attaches to the top ofthe fuselage by means of the top fuselage module and two bolts at the back. The wings centre moduleis one of the major and most important additions to the design. It firmly holds the wings in place,keeps them aligned, and secures them to the fuselage. It is a repeatable and reliable assembly. Thissetup extends the wingspan from 1800mm to 1910mm, which increases the lift force. Moreover, theinterior of the wings centre module provides additional space and still gives access to the interior ofthe fuselage.
49
(a) (b)
(c) (d)
Figure 3.16: (a) Top fuselage module. (b) Assembled wing centre module. (c) Assembly attachment glued tothe wings. (d) Wings assembly.
The tail was also redesigned (Figure 3.17). The default tail has wooden supports in the bottom thatconsiderably increase drag. This structure was replaced with a tail centre module that connects tothe tail boom and clears the way under the horizontal stabilizer. The vertical stabilizer is supposedto be glued to the horizontal stabilizer, but the new design allows the vertical stabilizer to be boltedonto the tail centre module, thus allowing disassembly for easy storage and transportation. Thedefault design has elevator and rudder servos mounted on the tail boom close to the fuselage, withexternal wires and long rods. The new design incorporates servos embedded into the foam in bothstabilizers. This reduces the length of the rods that connect to the rudder and elevator. Moreover, theservo wires travel inside the tail boom instead of outside. This new design has a cleaner setup thathelps reduce drag, improves aerodynamics, allows for easy disassembly, and is more reliable withprotected wires and short rods.
50
(a) (b)
Figure 3.17: (a) Original EPP FPV tail. (b) Modified tail (Falcon I design).
The wiring inside the fuselage is more efficient in the new design as cables are embedded andsecured into the foam (Figure 3.18a). For ease of assembly, wires are kept in just one side of thefuselage.
Figure 3.18: Internal wiring of the fuselage.
The new design uses the volume available in the nose of the plane for payload. A plastic nose (Figure3.19b) was designed for integrating any payload that could fit in such volume. With this approach,it is possible to have multiple noses, one for each type of payload, depending on the needs. Thebase design of the plastic nose can be modified to carry any piece of payload, provided volumeand weight permits. The nose attaches to the fuselage via six alignment pins and two screws. Thissystem provides a solid and reliable attachment while making it easy to install and uninstall the nose.The default foam nose (Figure 3.19a) can be used when no payload is needed. Additionally, the new
51
design has the battery sliding into the fuselage from the front, thus the nose has to be taken out forinserting and removing the battery. Figure 3.20 shows a rendering cut of the interior of the fuselagewith the electronics installed.
(a) (b)
Figure 3.19: (a) Default foam nose. (b) Custom plastic nose for installing a USB camera board.
Figure 3.20: Rendering cut of the fuselage with internal components.
The main electronics (system board, power board, switches board, ODROID single board computer)was installed on a base board that slides into the upper half of the fuselage, while the battery packslides into the bottom half. Switches in the front allow easy access to the user, and connections tothe UAV system board from the top contribute to keeping a clean and organized wiring. The centreof gravity of the redesigned airframe was kept within the same range of the original.
The redesign of the airframe provides a more efficient use of volume in the fuselage, that allows forbetter allocation and organization of the components inside. Even though plastic parts are heavier
52
than foam, thanks to the reduction in weight in the electronics assembly and imaging payload, theoverall weight of the fully assembled plane was reduced compared to the previous design. All plasticparts were design aiming for light-weight and high structural strength.
Previous designs had payload mounted on the the wings, while wings are kept intact in this redesign,with no holes or modification except for the attachment piece on the inside face, thus preserving theiraerodynamic capabilities. All the electronics was moved inside the fuselage, bringing all heavycomponents closer to the centre of gravity and making easier to balance the plane.
The result is a lighter plane with an efficient use of space inside the fuselage. The balance betweenthe original foam and the added plastic parts makes up for a reliable plane that achieves repeatabilityand can be assembled in a short time. The clean wiring and reduced size electronics make it easier tounderstand, learn, and handle by the user. The plane has been tested in multiple successful missions,and the University of Alberta Aerial Robotics Group (UAARG) adopted this redesigned version fortheir 2017 competition, and plan to continue working on top of this model. Figures 3.21 and 3.22show the final assembly of the plane.
Figure 3.21: 3D model of the Falcon I full assembly.
(a) (b)
Figure 3.22: Falcon I full assembly on the ground (a) and during flight (b).
53
Chapter 4
Proof of Concept MultispectralImaging System
This chapter covers the development and experiments carried out with the initial prototype of amultispectral imaging system. The approach used in the development of the multispectral imagingsystem follows a sequence of incremental steps based on prototyping and validation before thesubsequent step. Several iterations of the system were produced throughout the project and aredetailed in this chapter and the next. The first section of this chapter describes the initial work donewith USB camera modules, which led to the use of these cameras in the development of subsequentmulti-camera imaging systems. The second section of this chapter details the development of a3-band multispectral imaging system based on USB camera modules.
4.1 USB Camera Modules
There is a large variety of USB camera modules in the market targeted to areas that range fromresearch to industrial use. The first steps in the process of working with USB cameras was doing thenecessary research about this technology and becoming familiar with its features and capabilities.The main features, as well as the advantages and disadvantages of this type of cameras are coveredin 2.1.5. In this same section it is stated that the interest is focused on USB2.0 cameras compatiblewith the UVC driver.
54
4.1.1 Selection and Evaluation
The first step was to acquire a prototyping camera that could be used to validate the informationfound in the literature, as well as test the camera and evaluate its performance. With this in mind afirst selection was made. Camera ELP-USB500W02M from China based manufacturer ELP Ailipu
Technology Co.1 was selected as the first prototyping device. Table 4.1 shows the main features ofthis camera module.
Table 4.1: Features of the ELP-USB500W02M camera.
Manufacturer ELP Ailipu Technology Co.Part Number ELP-USB500W02MSensor Omnivision OV5640Sensor Type CMOSMaximum Resolution 2,592x1,944 pixelsPixel Size 1.4 µm
Image Area 3,629 x 2,722 µm
Shutter RollingPower Consumption ∼200 mA @ 5 VDCModule Size 38 x 38 mm
This camera was selected mainly because it provides a variety of settings that can be used to testmultiple configurations. It provides 10 different resolution configurations, both raw (YUYV) andcompressed (MJPG) output formats, and a variety of settings like brightness, contrast, hue, satura-tion, gamma, exposure, etc. In addition to its low cost, it is an excellent choice for performing allthe initial testing and evaluation. Figure 4.1 shows a picture of the ELP-USB500W02M.
Figure 4.1: USB camera module ELP-USB500W02M from ELP Ailipu Technology Co.
1http://www.elpcctv.com/
55
4.1.2 Software Development
Software development with this camera was done in a Linux environment and based on the use of thenative UVC driver (section 2.1.5.1) through the Video for Linux Two (V4L2) API (section 2.1.5.2).This type of camera can be compared to webcams in so much as components and interface. Thanksto the UVC driver the camera works as a plug-and-play device. It is possible to connect the camerato a computer via USB and open it using a video capture program (e.g. Cheese, Guvcview, VLC).For example, Guvcview2 will provide a Graphical User Interface (GUI) with a first window showinglive streaming video captured from the camera, and a second window with the settings accessible inthe camera.
Further interest was in finding a way to programmatically control the settings of the camera withouthaving to open a GUI. v4l2-ctl is a tool to access V4L2 drivers from the command line. This tool ispart of the v4l-utils library, which is a bundle of packages for handling media devices. Using v4l2-ctlis possible to obtain information from a connected camera and control its settings. Information suchas a list of all the settings with details of the type of field (e.g. int, bool), the range of accepted values,and the default value can be read from the camera. It is also possible to obtain a list of the possiblevideo output formats supported by the camera, as well as the combination of resolution and framerate options for each. Tables 4.2 and 4.3 list the camera settings and output formats respectively forthe ELP-USB500W02M.
Table 4.2: Settings of the ELP-USB500W02M camera.
Setting Type Min Max Step Default
Brightness integer 0 15 1 8Contrast integer 0 15 1 8Hue integer -10 10 1 0Saturation integer 0 15 1 7Sharpness integer 0 15 1 6Gamma integer 1 10 1 7White Balance Temperature integer 2,800 6,500 1 2,800Backlight Compensation integer 0 1 1 0PowerLine Freq. (Anti Flicker) menu 0 2 - 2Focus integer 0 21 1 16Exposure Auto menu 0 3 - 3Exposure integer 4 5,000 1 625
An useful feature of camera is the automatic exposure setting (Exposure Auto), where the camera
2http://guvcview.sourceforge.net/
56
takes care of adjusting the exposure and some internal gain setting in order to keep frames with theright amount of light.
Table 4.3: Frame Rates for the ELP-USB500W02M camera based on output video format and resolution.
ResolutionFrame Rate Frame Rate
@ MJPG @ YUYV
2592 x 1944 15 3
2048 x 1536 15 3
1920 x 1080 15 3
1600 x 1200 15 3
1280 x 1024 30 7.5
1280 x 720 30 7.5
1024 x 768 30 15
800 x 600 30 30
640 x 480 30 30
320 x 240 30 30
A second software tool was needed for capturing and saving frames from the camera. Several pro-grams were tested, including FFmpeg3, fswebcam4, Guvcview, uvccapture, VLC5, and GStreamer6
as command line tools. OpenCV7 was also tested as it is a popular library of programming functionsmainly aimed at real-time computer vision. Initial development was done using OpenCV, but someproblems were encountered when trying to configure the camera and synchronize processes.
With all of the command line tools being able to select an output format and resolution from thecamera, capture frames, and save them as image files at a certain rate, the decisive factor for a finalselection was performance.
GStreamer was selected as the best software tool because of its speed, low processing load, easeof use, and flexibility. But more importantly, as covered in section 2.1.5.3, GStreamer has theability to interface with multiple cameras simultaneously, which was a feature of interest for furtherdevelopment along this project.
A series of shell scripts were assembled using v4l2-ctl for configuring the camera and Gstreamer forcapturing frames.
3https://www.ffmpeg.org/4http://www.sanslogic.co.uk/fswebcam/5https://www.videolan.org/vlc/index.html6https://gstreamer.freedesktop.org/7https://opencv.org/
57
4.1.3 Lens Testing - Modulation Transfer Function
A group of lenses with different focal lengths were tested to find the resolution efficiency for each.Due to the lack of a part number to identify them, an easy ID number was assigned to each. A firstset of lenses (Lens ID 1-5) was acquired from the same China based manufacturer ELP Ailipu Tech-
nology Co. Lenses 1-4 did not have any pixel count specification whereas lens 5 was advertised asa megapixel lens rated for 5MP. Finally, one more lens (Lens ID 6) was acquired from the Germanybased manufacturer The Imaging Source8. Lens 6 was also advertised as a megapixel lens rated for5MP.
Lenses were tested following the ISO 12233 slanted-edge MTF method covered in section 2.1.4.3.The target shown in the same section (Figure 2.11) was printed with dimensions of 60x60 cm on amate white sheet with 300ppi quality using a laser printer. The ELP-USB500W02M camera runningat full resolution (2592x1944 pixels) with RAW (YUYV) output was used for taking the pictures.The camera was focused manually to the best focus possible. Pictures were taken, visually checkedfor focus, and processed using a software implementation of the ISO 12233 slanted-edge MTFmethod. The software was developed by Burns in the form of a MATLAB source code packagecalled sfrmat3: SFR analysis for digital cameras and scanners [89]. Each image was processed atleast twice and using different regions of interest in order to validate the results. As expected, themethod is robust and stable, giving consistent results through repeated processing.
The distance from the camera to the target was set so that the target would fill the FOV. Table 4.4shows the results of the tests. Lenses 1-4 showed a low resolution efficiency, with the lowest at 32%for lens 1, and the highest at 42% for lens 2. Lenses 5 and 6 showed a higher resolution efficiency,at 71% and 77% respectively.
Table 4.4: Results of different M12 lenses tested using the slanted-edge MTF method.
LensID
FocalLength
DistanceSamplingEfficiency
Vert.
SamplingEfficiency
Horiz.
ResolutionEfficiency
EffectiveResolution
(mm) (cm) (%) (%) (%) (MP)
1 3.6 77 59 54 32 1.59
2 6.0 115 64 66 42 2.11
3 8.0 158 63 58 37 1.83
4 12.0 261 61 60 37 1.83
5 6.0 120 71 100 71 3.49
6 6.0 120 79 97 77 3.83
8https://www.theimagingsource.com/
58
Figures 4.2 shows the resulting MTF plots of lens 5 for the horizontal and vertical components.In both cases it can be seen that plots present an increase in sharpness, evidenced by the positivebump that takes the MTF above 1.0 before dropping. This effect is more evident in the horizontalcomponent.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1Frequency (cycles/pixel)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
MTF
Horizontal (100%)Vertical (71%)
Figure 4.2: MTF plot for lens 5 (6mm focal length).
4.2 3-Band Multispectral Imaging System
A multispectral imaging system was developed using USB camera modules. Most basic multispec-tral imaging systems for vegetation and terrain analysis use 3-4 monochromatic cameras coveringthe Landsat Thematic Mapper bands TM2 (green), TM3 (red), and TM4 (NIR) as the most importantbands, and adding the TM1 (blue-green) or the narrow red-edge bands as complementary.
Due to space constraints, a 3-band multispectral imaging system was designed, developed, installedon the Falcon I UAV (section 3.4) and flown. The system is based on three USB camera modulesand optical filters. The cameras are connected to an ODROID SBC (section 3.2) via USB whichtakes care of controlling the cameras and saving frames into its local memory. The ODROID wasalso connected to the telemetry output of the autopilot board (section 3.1) to record avionics datastamps relevant for the images. Figure 4.3 shows the block diagrams of a camera subsystem and thecomplete imaging system.
59
(a) (b)
Figure 4.3: Simple block diagrams of a single camera (a), and the 3-band multispectral imaging system (b).
4.2.1 Cameras and Lenses
In section 2.2.3 a list of popular lightweight multispectral imaging systems currently in the marketwas presented. Based on an analysis of these systems, it was concluded that a resolution of 1.2MP (1280x960 pixels) was appropriate and reasonable for achieving a good balance between imagedetail, directly translated into GSD, and amount of data. It is important to keep in mind that USB 2.0has a theoretical maximum bandwidth of 480 Mbit/s or 60 MByte/s, and the more pixels a camerahas, the more data will travel via the USB connection. The system must be able to capture imagesfrom all cameras simultaneously, therefore care must be taken in making sure not to exceed the USB2.0 capacity. Camera ELP-USB130W01MT-B/W from ELP Ailipu Technology Co. was selected.Tables 4.5, 4.6, and 4.7 present the detailed features, settings, and output formats of this camera.
Table 4.5: Features of the ELP-USB130W01MT-B/W camera.
Manufacturer ELP Ailipu Technology Co.Part Number ELP-USB130W01MT-B/WSensor ON Semiconductor AR0130CSSensor Type CMOSMaximum Resolution 1,280x960 pixelsPixel Size 3.75 µm
Image Area 4,800 x 3,600 µm
Shutter RollingPower Consumption ∼160 mA @ 5 VDCModule Size 38 x 38 mm
60
Table 4.6: Settings of the ELP-USB130W01MT-B/W camera.
Setting Type Min Max Step Default
Brightness integer -64 64 1 0Contrast integer 0 95 1 32Hue integer -2,000 2,000 1 0Saturation integer 0 128 1 0Sharpness integer 1 7 1 2Gamma integer 100 300 1 0White Balance Temperature Auto bool 0 1 - 1White Balance Temperature integer 2,800 6,500 1 4,600Backlight Compensation integer 0 3 1 1Gain integer 0 100 1 0PowerLine Freq. (Anti Flicker) menu 0 2 - 1Exposure Auto menu 0 3 - 3Exposure integer 1 5,000 1 625Exposure Priority bool 0 1 - 0
Table 4.7: Frame Rates for the ELP-USB130W01MT-B/W camera based on output video format and resolution.
ResolutionFrame Rate Frame Rate
@ MJPG @ YUYV
1280 x 960 30/15 9
1280 x 720 30/15 10
800 x 600 30/15 20
640 x 480 30/15 30
320 x 240 30/15 30
With 1280x960 pixels and a pixel size of 3.75 µm, the image sensor in this camera has the sameresolution capabilities than the Sentek GEMS, Sentera QUAD, and Parrot Sequoia cameras (section2.2.3). The major difference is that the ELP-USB130W01MT-B/W has a rolling shutter insteadof a global shutter. This poses a challenge as imaging issues like skew or blur could be presentin the final images under certain conditions. Another challenge is synchronizing multiple ELP-USB130W01MT-B/W cameras. Since there is no way of doing this via hardware, synchronizationwould have to be done through software.
It is important to take into consideration that this type of CMOS image sensor have a high quantumefficiency (measure of sensitivity of the photodiodes) in the visible wavelengths range (above 70%),
61
but it decays at higher wavelengths (Figure 4.4). Therefore, the NIR range will have a lower quantumefficiency that will have to be compensated with a longer exposure time or a higher gain.
350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150Wavelength (nm)
0
10
20
30
40
50
60
70
80
90
Qua
ntum
Effi
cien
cy (%
)
Figure 4.4: Quantum Efficiency of the Aptina / ON Semiconductor AR0130CS image sensor.
Lens 5 (6 mm focal length, 5 MP) was selected for the imaging system. The higher resolutionefficiency (71%) gives better imaging performance. Going with 6 mm focal length results in a FOVof 43.60° x 33.39° and a GSD of 7.5 cm @ 400 ft (~122m). Compared to values presented in table2.4 (section 2.2.3) for commercial cameras, this configuration is close to the ones adopted by theSentek GEMS, Sentera QUAD, and Parrot Sequoia cameras. The average between the focal lengthsof these cameras gives 5.56 mm, close to the 6 mm focal length selected for our system. Thisselection offers a good balance between FOV and GSD.
4.2.2 Optical Filters
Filters for the imaging system were selected following the general convention of bands used incommercial multispectral cameras (section 2.2.3), which in turn is oriented towards the calculationof vegetation indices. Given that the system has three cameras, the green, red, and NIR bands wereselected. Filters were acquired from Andover Corporation 9. Table 4.8 and Figure 4.5 present thetechnical details and spectral response of the filters. Figure 4.6 shows a comparison of the bandschosen for this system and the bands used in other commercial multispectral imaging solutions.
9https://www.andovercorp.com/
62
Table 4.8: Technical details of the optical filters.
Parameter Green BandFilter
Red BandFilter
NIR BandFilter
Center Wavelength (nm) 550 +10/-0 650 +10/-0 800 +10/-0Bandwidth (nm) 40 +8/-8 40 +8/-8 40 +8/-8Transmission (%) 50 50 50Size (mm) 12.5 +0/- 0.25 12.5 +0/- 0.25 12.5 +0/- 0.25Thickness (mm) 5.9 5.9 5.9Refractive Index 2.05 2.05 2.05
400 500 600 700 800 900 1000 1100Wavelength (nm)
0
10
20
30
40
50
60
70
80
90
Spec
tral R
espo
nse
(%)
GREENREDNIR
Figure 4.5: Spectral response or transmissivity of the green, red, and NIR band-pass filters chosen for thesystem.
Figure 4.6: Comparison of chosen bands with bands used in other commercial multispectral cameras.
63
Filters came as discs of 12.5 mm of diameter. A small adapter was designed for holding the filterand fit on top of the lens as a cap (Figure 4.7). The adapter was 3D printed, the filter would fit tightlyinside, and the whole cap would sit on top of the lens, secured by a screw.
Figure 4.7: From left to right; green, red, and NIR filters mounted on lens cap adapters.
Lens 5 comes with an IR cut filter glued to the back of the lens (Figure 4.8). This IR cut filter wasremoved from the lens where the NIR filter was installed, so as to be able to capture the desiredband. The filter is a small glass disc with glue in three points. The filter was removed using a smallsharp knife to cut through the points of glue.
Figure 4.8: Back of a lens with IR cut filter (left), and after IR cut filter removal (right).
4.2.3 Software Development
As previously mentioned in section 4.1.2, a series of shell scripts were created in order to interfacewith the camera. The scripts are based on v4l2-ctl for configuring the camera and Gstreamer forcapturing frames. The v4l2-ctl configuration sets the values of the camera settings shown in table4.6. Default settings were used with automatic exposure enabled.
After configuration, three identical pipelines were set up to run in parallel using Gstreamer. Eachpipeline starts by opening a multimedia source, which in this case is a V4L2 source (camera). The
64
second step is to select the video output option. In this case, raw video (YUYV) with 1280x960resolution and 9 fps frame rate was selected. The next step is to set the rate at which frames aregoing to be extracted from the source. A 1 fps frame rate was selected, which means that the sourceof 9 fps is downsampled to 1 fps. In other words, one frame is kept and 8 frames are skippedevery second. The reason for selecting this frame rate depends on the conditions of the flights andis explained in more detailed in section 4.2.5. Final steps are to convert and encode the frame as aPNG file and save it in local memory. By using a multifilesink sink the pipeline will run in an endlessloop. Figure 4.9 shows the diagram os the Gstreamer application.
Figure 4.9: Diagram of the Gstreamer pipeline application for three ELP-USB130W01MT-B/W cameras.
While the above was the optimal configuration defined for the system, the software package wasdesigned to keep settings in external configuration files that are given as input arguments to the mainscript. This way it is possible to define multiple configuration files with settings that are easy tochange, thus making the overall software flexible.
Codecs JPEG, TIFF, and RAW were also tested for the encoder section. JPEG is a compressedformat that leads to a faster processing and smaller file size, but by being compressed some infor-mation is lost. TIFF and RAW and complex lossless formats that keep all the information but takelonger to process and lead to larger file sizes. PNG is a simple format that doesn’t have most of thenative features of TIFF or RAW while still being lossless and maintaining the image data. PNG wasselected as the optimal format for being lossless and resulting in relatively small file sizes.
65
4.2.4 UAV Setup
In section 3.4.1 the interchangeable payload nose was introduced. This concept gives flexibilityto the airframe design as multiple noses can be designed using the base model. The purpose ofthe noses is to hold the payload systems, which in this case is the cameras. A custom nose wasdesigned to hold the three ELP-USB130W01MT-B/W cameras with filters attached. Given thespace available in the nose base model, the cameras were arranged in three different levels (Figure4.10). This arrangement introduces a distance difference of 13 mm between the sensor planes ofeach camera sensor, which is negligible given the intended range of working distances. The upsideof this arrangement is that the optical centre of the cameras is closer to each other (29 mm), reducingthe amount of misalignment between the cameras. Cameras are secured to the nose using screwsand connected to the onboard ODROID SBC via USB cables. The figure shows how the locationof the filters was assigned. The bottom part of the nose includes a transparent acrylic window thatprotects the cameras and preserves aerodynamics.
(a) (b)
Figure 4.10: 3D model of custom plane nose with 3 cameras. (a) Isometric view. (b) Side cross-section view.
(a) (b)
Figure 4.11: (a) Custom plane nose with 3 cameras and filters. (b) Nose installed in the plane.
66
The 3D printed custom nose with the 3 cameras, lenses, filters, and all the hardware has a totalweight of 154 grams. This doesn’t include the USB cables and ODROID computer that is necessaryfor interfacing with the cameras. Nonetheless, it is a relatively lightweight setup considering the useof 3 USB camera modules. The addition of this payload to the UAV does not increase the overallweight considerably and has a minor effect in the overall center of gravity.
4.2.5 Flight Campaigns
A series of flight campaigns were performed in order to test the system. Flights were performed atBremner Field, a grass field in the North Saskatchewan River Valley, north of Sherwood Park, inAlberta, Canada. It is located in coordinates 53.6380961°, -113.2895888°, with an altitude of 610m above sea level. This field offers north-south, east-west, and northwest-southeast runways withclear and unobstructed approaches from all directions. Bremner Field is managed by the EdmontonRadio Control Society (ERCS)10. Figure 4.12 shows a map of Bremner Field with the delimited areaallowed for flight operations.
Figure 4.12: Map of Bremner Field indicating the delimited area for flight operations.
There was a total of four flights done with the 3-band multispectral imaging system on board of theFalcon I UAV. Table 4.9 presents the details of each flight. Flights 3 and 4 were divided in 2 and
10https://www.ercs.ab.ca/
67
3 segments respectively with different altitudes each. Actual speeds and altitudes are reported asaverages in the table.
Table 4.9: Flight campaigns details.
Flight Date StartTime
Duration(mm:ss)
Avg. GroundSpeed (m/s)
Avg. Altitude(m)
ImageSets
1 2017-AUG-03 12:13:30 03:08 17 60 1882 2017-AUG-03 12:32:20 08:13 17 100 4933 2017-AUG-24 14:36:46 15:27 21 50 / 90 9274 2017-SEP-15 12:34:13 19:49 18 60 / 80 / 100 1189
Flights involved manual take-off and landing, and a mix of assisted and autonomous flying. Assistedflying allows the pilot to control the direction and speed of the UAV but it limits the amount of roll,pitch, and yaw. Autonomous flight gives total control of the UAV to the onboard autopilot system,which executes a predefined flight plan. An operator in charge of the GCS monitors the behavior ofthe UAV at all times and manages the flight plan when in autonomous mode.
The imaging scripts were activated manually via terminal in the ODROID before taking off, andterminated in the same way after landing. Given the ground speed and altitude of the UAV, theimage capture frequency was set to 1 frame per second. At this capture frequency there is a 33.33%overlap between two consecutive frames flying at 50 m AGL and 20 m/s ground speed, and a 66.67%overlap for 100 m AGL and the same speed. Cameras were used at maximum resolution (1280x960pixels) and focus was set for a range of 50 - 100 meters, using the principle of hyperfocal distance,which is the the closest focusing distance that allows objects at infinity to be acceptably sharp. Underthese conditions, images did not show any visible skew or blur during normal flight (take-off andlanding excluded).
4.2.6 Image Alignment
Due to the use of multiple cameras and the fact that its optical centers are misaligned, the resultingimages will also be misaligned. Alignment of the images is necessary in order to operate on themand obtain useful information like vegetation indices. In the UAV setup presented above the camerasare horizontally aligned (horizontal being the direction perpendicular to the direction of flight) butvertically misaligned (vertical being the direction of flight), with a distance between optical centersof 29 mm. This means that alignment of the images can be done operating only on one axis.
A MATLAB program was developed in order to run an alignment algorithm over the images. Thealgorithm takes the image from the middle camera (with the NIR filter) as the reference and alignsthe other two (red and green) with respect to it. The algorithm is based on correlation between theimages, where one image is the reference and the other is the subject.
68
The algorithm is based on the assumption that there is a single maxima in the pixel-by-pixel correla-tion between two misaligned images as one of the images is translated a certain amount of pixels ina certain direction. Since the images in this case are misaligned only in one axis, and the directionof misalignment is known, the problem is simplified as the algorithm runs only in one axis and onedirection (1D translation). Figure 4.13 shows the correlation plot of two images as one them is keptfixed and the other is displaced. Displacement was done with 1 pixel increments with a range of50 pixels. The maxima for this example occurs at 22 pixels, with a correlation of 0.84312. At thispoint, the two images have the highest correlation between them. If the pixel offset, 22 pixels inthis case, is cropped from both images, one at the top and the other at the bottom, the result is twoaligned images. The images used for this test were taken at approximately 60 meters away from thetarget. At larger working distances the misalignment is less and the offset in pixels will be less.
0 10 20 30 40 50Displacement (pixels)
0.7
0.75
0.8
0.85
0.9
Cor
rela
tion
(22, 0.84312)
correlationmaxima
Figure 4.13: Correlation plot of alignment algorithm indicating maxima.
The process for aligning a set of three images uses the same principle. Two images are alignedfirst to produce a first aligned subset, then the third image is aligned with respect to this subset toproduce the final set. In this case, the middle image and one of the side images is aligned first.After alignment and cropping, the new middle image is used as reference to align the other sideimage. Then, the offset from the first pair and the second pair is compared, taking the greater oneand adjusting in order to make sure all three images have the same size. The final images will endup with a smaller height due to the cropping (Figure 4.14). The areas of the images, top and bottom,that are not present in all three images are discarded, leaving three images accurately aligned.
69
Figure 4.14: Example of pixel offset and cropping of images for alignment.
Figure 4.15 shows the example of three images (red, green, and NIR bands) superposed before andafter alignment. In this example the misalignment between red and NIR band images was 19 pixel,and between NIR and green band images was 21 pixels.
Figure 4.15: Superposition of red, green, and NIR band images before and after alignment.
70
Figure 4.16 shows a closer comparison of the superposed images before and after alignment.
(a) (b)
Figure 4.16: Result of image alignment algorithm. (a) Superposition before alignment. (b) Superposition afteralignment.
While this method might seem simple compared to more advanced algorithms [90], it gives accurateresults and takes less processing time. The algorithm was improved by making it exit the correlationcomparison loop as soon as the maxima was found. It takes the algorithm an average of 1.79 secondsto align a set of three images taken at a working distance of 80 meters, where the misalignment is inaverage 19 pixels.
The resulting images after alignment end up being shorter in height as only common areas of theimages are kept, the rest is removed. The final superposed image is clear and well defined. Sincelens distortion was not corrected the alignment of pixels is not perfect, and some offset can be seen.The offset goes from zero at the center of the image and grows moving outwards. It is possible tosee misaligned areas at the edges and corners of the superposed image. Nonetheless, the results arefast and accurate enough for basic pixel level operations in the images.
4.2.7 Vegetation Indices
After capturing images of the same scene using different spectral bands, and aligning the images, itwas possible to operate over the images and calculate vegetation indices. Although there are severalvegetation indices, the most widely used is the Normalized Difference Vegetation Index (NDVI),introduced in section 2.3.1. The NDVI is a simple numerical indicator used for analyzing remotesensing imagery and determine whether the target contains live green vegetation or not. NDVIimages were calculated from NIR and red band images using equation 2.11 from section 2.3.1 on apixel by pixel basis. Resulting images have a [-1,1] range and are false colored in order to obtain aclear vegetation map.
71
Figure 4.17 shows a test performed with the 3-band multispectral imaging system presented in thiswork, using six leaves of the same family, each with different levels of stress, going from a perfectlyhealthy leaf to a dead leaf. The experiment was done using natural sunlight as the incident source.
(a) (b)
Figure 4.17: Example NDVI test of leaves with different levels of stress. (a) RGB image. (b) False coloredNDVI image.
Classification within the range of NDVI values is important in order to correctly identify the areasof interest in the image. Barren areas of rock, sand, or snow have a low NDVI value (0.1 andbelow). Sparse vegetation like shrubs and grasslands show a moderate NDVI value (approximately0.2 to 0.5). High NDVI values (approximately 0.6 to 0.9) correspond to dense vegetation commonlyfound in temperate and tropical forests or crops at their peak growth stage. Bare soil is representedwith NDVI values that are closer to 0, and water bodies are represented with negative NDVI values[91]. Although these are general thresholds, in practice it is usually necessary to perform some realsample calibration to adjust the ranges for up to ± 0.1.
In the experiment with the leaves, the color scale was adjusted for moderately healthy vegetation tostart at an NDVI value of 0.25. It is evident in the image, as it is compared to the equivalent RGBimage, that healthy leaves can be clearly differentiated from highly stressed (dry and dead) leaves.Furthermore, clear levels can be identified in the three healthy leaves on the NDVI image.
All the NDVI processing algorithm (calculation and false coloring) was done in MATLAB andpacked in a single processing script. The same script was later used for processing the images ofthe flights after alignment. Figure 4.18 shows samples of NIR band, red band, and NDVI imagesof pictures captured from the UAV at Bremner Field. In this field, the main runways have livegreen grass that is regularly watered and in maintained conditions, whereas the surrounding sectionsof the runways have grass and bushes that live in natural conditions. The surrounding sectionsexhibit less green vegetation and higher levels of dryness to the eye. NDVI images show a cleardifferentiation between the maintained and apparently healthy live grass in one of the main runwaysand the surrounding not maintained grass.
72
Figure 4.18: Sample pictures of a flights at Bremner Field showing the NIR and Red band images and theNDVI resulting image.
73
Chapter 5
12-Band Multispectral ImagingPlatform
This chapter covers the design, development, and experiments carried out with the more advancedprototype of a 12-Band multispectral imaging system. The first section introduces the concept andkey aspects of the system. The second section covers the USB camera modules chosen for the sys-tem, explaining the factors and reasons for its selection. Subsequently, the custom camera bridgeboard is introduced, covering the integrated and configurable USB HUB and automatic trigger cir-cuits. Finally, results from testing are presented.
5.1 System Concept
The proposed system will follow the approach presented in the previous section. In this revision,the system is based on an array of up to 12 cameras with global shutter and synchronized trigger.The cameras continue to have a USB 2.0 interface, and will be connected to a custom designed PCBthat will include USB HUBs to integrate USB routes and minimize cabling. The USB routes willbe configurable to a degree to allow for grouping cameras depending on configuration and USBbandwidth requirements. The custom camera bridge board will also include and triggering circuitfor the cameras based on a microcontroller unit (MCU).
The goal is to create and test an integrated multi-camera system based on USB camera modules thatis flexible and configurable, so that it can be used with optical filters as a development platform formultispectral imaging applications.
74
5.2 USB Camera Modules
The imaging system was designed to work with the DMM 42BUC03-ML USB USB 2.0 board-level camera from the manufacturer The Imaging Source, LLC1. Selection of this camera was madebased on a balance between size, cost, and capabilities. The requirements were set for a camerawith global shutter and trigger as the main specifications. The usual maximum resolution for globalshutter sensors is 1.2 Megapixels. The use of S-Mount lenses and no packaging or enclosure reducesthe costs, and the size was selected as the minimum possible to allow for miniaturization and lightweight. The DMM 42BUC03-ML (Figure 5.1) is a small 30 mm x 30 mm monochromatic board-level camera module with S-Mount (M12) lens based on the ON Semiconductor AR0134CS CMOSimage sensor. This sensor comes with a global shutter [92], which is ideal for imaging a target orscene in movement as it does not suffer from imaging issues associated with rolling shutter such asskew and wobble [93]. Table 5.1 summarizes the technical specifications of this camera.
Figure 5.1: DMM 42BUC03-ML camera module from The Imaging Source, including S-Mount lens holderand lens.
Table 5.1: Specifications of the DMM 42BUC03-ML camera.
Manufacturer The Imaging SourcePart Number DMM 42BUC03-MLSensor ON Semiconductor AR0134CSSensor Type CMOSMaximum Resolution 1,280 x 960 pixelsPixel Size 3.75 µmImage Area 4,800 x 3,600 µmShutter GlobalPower Consumption ∼250 mA @ 5 VDCModule Size 30 x 30 mm
1https://www.theimagingsource.com/
75
The ON Semiconductor AR0134CS has 1.2 Megapixels (1,280 x 960 pixels) and a pixel size of3.75 µm, which makes it equivalent to the image sensors used in the Parrot Sequoia, MicaSenseRedEdge-M, Sentera QUAD, and Sentek Systems GEMS Multispectral Sensor. The quantum effi-ciency (measure of the sensitivity of the photodiodes) of this sensor (Figure 5.2) is also very similarto the ones reported for the commercial cameras.
350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150Wavelength (nm)
0
10
20
30
40
50
60
70
80
Qua
ntum
Effi
cien
cy (%
)
Figure 5.2: Quantum efficiency of the ON Semiconductor AR0134CS image sensor.
The DMM 42BUC03-ML camera also features a working mode based on an external trigger input.When a trigger signal, defined as a positive pulse of at least 10 microseconds, is received by thecamera in the trigger input pin, the camera will start the exposure and readout process of a singleframe (Figure 5.3). The length of the exposure can be set by software. The duration of the imagereadout is the reciprocal of the current frame rate. After image readout is finished, the camera is ableto accept another trigger. It is possible use this feature to send the same trigger signal to multiplecameras and have them synchronized. The trigger input on the camera is opto-coupled and admitspositive trigger pulses with any amplitude between 3.3V and 12V.
Figure 5.3: Timing diagram of the trigger, exposure, and image readout processes.
This camera has a monochromatic raw output video format (Y800) and can operate at three different
76
resolution options, with several frame rate options for each resolution (Table 5.2). When selectinga lower resolution, contrary to pixel binning [94], the DMM 42BUC03-ML will only read the cor-responding amount of pixels referenced to the top left corner of the sensor. Therefore, lowering theresolution reduces image data and USB bandwidth requirements, which makes possible for morecameras to share the USB resource.
Table 5.2: Video formats and frame rates available in the DMM 42BUC03-ML camera.
ResolutionFrame Rate
@ GREY (Y800)
1280 x 960 30 / 25 / 15 / 10
640 x 480 60 / 30 / 25 / 15
320 x 240 60 / 30 / 25 / 15
The USB interface in this camera is compatible with the USB Video Class (UVC) standard driver.This camera offers a range of settings that are accessible and configurable via software (Table 5.3).The “Privacy” value is the trigger enable/disable setting. When disabled the camera is in free runningmode. When enabled the camera is in trigger mode.
Table 5.3: Settings of the DMM 42BUC03-ML camera.
Setting Type Min Max Step Default
Brightness integer 0 255 1 12Gain integer 34 255 1 36Exposure integer 1 300,000 1 127Privacy bool 0 1 - 0
5.2.1 Optical Filters
The use of S-Mount (M12) lenses in these camera modules allow the use of optical filters with thesame approach presented in section 4.2.2. The 12.5 mm of diameter filter discs are easily installedin the lenses using a small 3D printed adapter piece that holds the filter and fits in front of the lens,secured by a screw (Figure 5.4).
5.3 Camera Bridge Board
The camera bridge board (Figure 5.5) acts as a motherboard for the camera modules and is themain integration unit of the system. Cameras connect directly to the bridge board, which routes all
77
Figure 5.4: Camera board with S-Mount (M12) lens and optical filter installed using the 3D printed adapter.
USB communications via USB HUBs into selectable upstream USB ports. This board includes astep-down voltage regulator that provides power to all components including the cameras. It alsointegrates a microcontroller unit (MCU) that generates the pulses used for triggering the cameras.
Figure 5.5: Diagram of the basic connections between the camera array and the bridge board. The bridgeboard routes all USB communications into selectable upstream ports and sends trigger signals to the cameras.
The goal was to design a compact system, and avoid the use of USB cables as they tend to be bulkyand take up considerable space and mass. The bridge board was designed to have straight mini USBplug connectors that match the mini USB receptacle connectors found in the DMM 42BUC03-MLcamera modules. The trigger input pins of the camera modules are found in a picoblade connector.The bridge board has cutouts underneath of each camera location, and individual trigger connectorson the other side of the board. Mounting holes are also included for securing the cameras to thebridge board using standoffs.
The bridge board Printed Circuit Board (PCB) was designed using the PCB CAD tool Altium De-signer. Schematics and Bill-of-Materials of the camera bridge board can be found in Appendices Band C. Figure 5.6 shows a layout view of the board with the top and bottom layers.
78
Figure 5.6: Altium Designer layout view of the bridge board with top layer (red), bottom layer (blue), andsilkscreen (yellow).
The PCB is a 4-layer board. External layers are used for routing signals and internal layers areplanes. Layer 1 (top) contains most of the components and USB routing. Layer 2 is a ground planethat provides reference for the USB differential pairs so as to keep the desired impedance. Layer 3is a power plane, and layer 4 (bottom) contains the trigger connectors and additional routing. Thebridge board alone weighs 95 grams, and is 150 mm long by 138 mm wide. Figure 5.7 shows aCAD 3D model of the bridge board.
Figure 5.7: CAD 3D model of the bridge board.
79
It is important to note that the bridge board itself is a USB 2.0 multi-camera platform that can bemodified and adapted to work with a wide range of cameras. While this board was designed for theuse of the DMM 42BUC03-ML camera modules, it is perfectly possible to connect other cameras.The Imaging Source provides other USB 2.0 board-level camera modules with different features butkeeping the same form factor. Therefore it is possible to replace the selected DMM 42BUC03-MLcamera module with other cameras from the same manufacturer and still be able to utilize the directconnection to the mini USB connector, mounting holes, and trigger connection. Alternatively, it isalso possible to use any other camera module as long as it communicates via USB 2.0. Additionalcables might be needed to complete the USB and trigger connections between the cameras and thebridge board, and extra work might be required for attaching the cameras.
The bridge board provides all necessary electronics for the connection of multiple USB 2.0 camerasto different upstream USB ports via configurable USB HUBs. It also provides a configurable triggersystem handled by a reprogrammable MCU. The following sections cover in detail specific featuresof the bridge board.
5.3.1 USB HUB
A configurable USB 2.0 HUB in the board allows for the connection of up to 12 USB 2.0 cameramodules (Figure 5.8). The cameras are organized in groups of four, going trough a first level of USBHUBs and then USB switches that offer two different routing options. The first option is a secondlevel USB HUB that connects all cameras to a master USB upstream port. The second option is toroute each group to its dedicated USB upstream port. The USB switches are controlled with userselectable jumpers. The flexibility of this approach allows the user to choose between isolating andmerging USB communications depending on the USB bandwidth requirements.
Figure 5.8: Diagram of the configurable USB HUB network. SW 1, SW 2, and SW 3 are 2-to-1 bidirectionalUSB 2.0 switches with inputs controlled by user selectable jumpers. USB 1, USB 2, USB 3, are the isolatedupstream ports, and USB M is the master upstream USB port.
80
By isolating a first layer HUB we end up with an independent group of 4 cameras routed to anupstream mini USB port (USB 1, 2, 3). A lower number of cameras reduces the load and bandwidthrequirements on a USB port. This setup is used for cameras that require a high USB bandwidth.Alternatively, when using cameras with a lower USB bandwidth requirement, it is possible to mergesome or all USB communication in the bridge board into a single USB port (USB M). This approachsimplifies the connection of the bridge board to an external computer.
5.3.2 Trigger
The bridge board includes a trigger system based on a microcontroller unit (MCU). The MCU isan 8-bit 16Mhz ATMEGA328P-AU compatible with the Arduino framework and development en-vironment [95, 96]. The MCU uses general purpose input/output (GPIO) pins to generate the pulsesthat trigger the cameras. A total of 6 GPIO pins from the same MCU port are used, and each pin con-nects to a group of 2 cameras. By having 6 independent trigger pins it is possible to define customtriggering sequences for each pair of cameras. Conversely, by having the 6 pins from the same MCUport it is possible to trigger all of them at the same time, thus allowing accurate synchronization ofall cameras. Trigger signals go through a jumper section in the board, allowing the user to break theconnection between the MCU and the cameras. By means of this it is possible to connect the MCUpins to a different location, and/or connect trigger signals from an external system. This approachbrings more flexibility to the design, making it easy to modify according to the requirements. Figure5.9 shows a diagram of the trigger system.
Figure 5.9: Diagram of the trigger system based on a microcontroller unit (MCU). Trigger signals are gener-ated via GPIOs. A UART connection via USB is provided for interactively interfacing with the MCU via serialcommands.
The MCU is also connected via UART to a UART-to-USB transceiver, which in turn connects to aUSB connector. This addition provides an easy way to flash the MCU as long as the correspondingbootloader is on. More importantly, this connection provides a way to communicate interactivelywith the MCU via USB serial. A set of commands has been defined in the default MCU program tomake it possible to change some settings of the trigger system on the go, without needing to modify
81
the program and re-flash it. Features like changing the period of the automatic trigger, start or stopthe automatic trigger, send trigger signals based on a command, are included in the MCU programand can be controlled via serial commands. The ability to have multiple functions and modes ofoperation of the trigger system integrated in the same program makes it easy, flexible, and versatile.
5.4 Software
Similar to the previous imaging system (Chapter 4), a software package was developed for interfac-ing with the imaging system. The core of the software remained in the two elements: The Video forLinux Two (V4L2) API, used to interface with the lower-level UVC driver, and GStreamer, used forsimultaneous interface with multiple cameras.
Similarly, the software package is based on a collection of shell scripts that take care of configuringthe cameras and building the acquisition, processing, and storing pipelines for each camera. Thescript runs according to settings specified in a configuration file that is given to the script as anargument. A series of external configuration files keep different variants of the system and allow forcustomization of the program.
The trigger system MCU runs a program written in C++. The program sends simultaneous pulsesto all cameras periodically. The triggering period is set to 1 second by default. This value is loadedfrom the MCU EEPROM. The program also has a serial terminal-like interface that accepts customdefined AT commands. The commands allow to stop and restart the periodic trigger, change thetriggering period (any new value is saved in the EEPROM), and send a single trigger.
5.5 Testing
An experiment was designed to test the system in the lab with different configurations of the cameras,and determine the effect in the USB traffic load, as well as the maximum number of cameras thatcan work simultaneously for a given configuration. The platform was tested with DMM 42BUC03-ML camera modules synchronized using the trigger system. The system was connected using asingle upstream USB port to an external computer running Linux. USB traffic was analyzed usinga sniffing tool called Wireshark2 [97]. Resolution, Frame rate, and the number of simultaneouscameras were the controlled variables. The total throughput in the USB 2.0 channel was monitored.Table 5.4 summarizes the scenarios that successfully worked with images captured simultaneouslyfrom all cameras and saved locally.
Images from the DMM 42BUC03-ML are transmitted over USB in chunks (packets) of 16,448bytes. First, the host (computer) sends a packet request, and then the peripheral (camera) sends the
2https://www.wireshark.org/
82
Table 5.4: Test results of USB traffic with different configurations. Cameras were connected to a single USBupstream port and triggered simultaneously. Data transferred based on a single frame from each camera.
Test # Resolution FrameRate
#Cam
DataTransferred
Throughput
(fps) (Bytes) (MB/s)
1 1280 x 960 30 1 1,238,530 38.982 1280 x 960 10 1 1,238,530 12.933 1280 x 960 10 2 2,477,060 25.904 1280 x 960 10 3 3,715,590 38.955 640 x 480 30 2 619,268 20.876 640 x 480 15 2 619,268 10.437 640 x 480 15 4 1,238,536 20.878 640 x 480 15 8 2,477,072 41.619 320 x 240 15 8 619,280 10.32
10 320 x 240 15 12 928,920 15.5111 320 x 240 30 12 928,920 29.44
response packet (includes overhead and payload). This process continues until the whole frame hasbeen transferred. The total data transferred reported in Table 5.4 takes into accounts all traffic, whichincludes request packets and response data packets. As mentioned before, the maximum throughputin practice for USB 2.0 is around 40 MB/s. Tests 1, 4, and 8 have working configurations at the limitof the USB bandwidth capacity. Adding an additional camera to these configurations would exceedthe USB 2.0 bandwidth and would result in dropped packets representing permanently lost fractionsof camera frames.
It is evident that decreasing the resolution leads to less data being transferred, and lowering theframe rate gives more time for data to transfer. Both of these conditions cause the throughput todecrease and allow the connection of more cameras. The resolution and frame rate settings definethe maximum number of cameras that can work at the same time.
Based on the results of these tests it is possible to estimate the maximum number of cameras thatcan work simultaneously with a certain configuration. The estimation uses the maximum practicalUSB bandwidth of the system and an estimate of the total throughput.
Max # Cameras =USB BW
T hroughput(5.1)
The USB BW is set to 40×106 Bytes/s, and the Throughput is given by
83
T hroughput =Data Frame
Trans f er Time(5.2)
Data Frame = SF× (Width×Height) (5.3)
Trans f er Time =1
Frame rate(5.4)
Since the Y800 format used by the cameras assigns one byte per pixel, Data Frame is estimated asthe total number of bytes required by the image, multiplied by a scaling factor (SF = 1.01) to accountfor the response packets overhead and request packets. This scaling factor was calculated from thesize of the request packets, response packets, and response packets overhead. Each request packetis 64 bytes, and each response packet is 16,448 bytes out of which 100 bytes are overhead and therest is payload. The scaling factor is the ratio of the total number of bytes exchanged via USB tothe actual number of payload bytes (16,512 / 16,348 = 1.01). The transfer time is equivalent to thereadout time, which is the reciprocal of the frame rate. After substituting (5.2), (5.3), and (5.4) in(5.1) we get
Max #
Cameras =USB BW
SF×Width×Height×Frame rate(5.5)
The integer part of this number gives an accurate estimation of the maximum number of camerasthat can work simultaneously for a certain resolution and frame rate.
The final system, shown in Figure 5.10, has been successfully tested with an ODROID C2 singleboard computer (SBC). Given the low CPU load of the software, it can run with no issues in manyrelatively cheap SBCs. The imaging board is 150 mm long by 138 mm wide, and 43 mm tallconsidering the camera modules and lenses. Although the size and focus point of the lenses willaffect the overall height. The board weights 323 grams with 12 DMM 42BUC03-ML cameras,standoffs, and fasteners.
84
Figure 5.10: Multispectral imaging system hardware with 12 cameras installed.
Finally, table 5.5 shows an extension of table 2.4 presented previously in section 2.2.3, with theaddition of the 12-band multispectral imaging system presented in this section for comparison.
Table 5.5: Comparison of common multispectral imaging systems with the system presented in this work.
Camera This Work Tetracam
ADC Micro
Tetracam
Mini-MCA4
Sentek
GEMS
Sentera
QUAD
Parrot
Sequoia
Dimensions (mm) 150 x 138 x 43 75 x 59 x 33 131 x 78 x 87 127 x 89 76 x 62 x 48 59 x 41 x 28
Weight (g) 323 90 600 320 170 72
Number of Bands 12 4 4 4 4 4
Focal Length (mm) 6.00 8.43 9.6 7.7 5.0 3.98
FOV (H° x V°) 43.6 x 33.4 37.6 x 28.7 38.2 x 30.9 34.6 x 26.3 51.2 x 39.6 61.9 x 48.5
Sensor size (mm) 4.8 x 3.6 6.55 x 4.92 6.66 x 5.32 4.8 x 3.6 4.8 x 3.6 4.8 x 3.6
Resolution (pixels) 1280 x 960 2048 x 1536 1280 x 1024 1280 x 960 1280 x 960 1280 x 960
Pixel size (µm) 3.75 3.12 5.2 3.75 3.75 3.75
GSD (cm) @120m 7.5 4.6 6.6 5.1 9.1 11.0
Price (USD) See Table 5.6 2,995.00 9,995.00 3,500.00 4,599.00 3,500.00
85
5.6 Costs
Tables 5.6 presents a list of general costs for different configurations of the multispectral imagingsystem hardware. Table 5.7 presents the costs per unit or set for the Bridge Board bare PCB andBOM based on order quantity. All prices are in US Dollars (USD).
Table 5.6: General costs of components for the multispectral imaging system hardware.
Components 3 bands @
1.2MP
8 bands @
0.3 MP
9 bands @
1.2MP
12 bands @
0.3MP
Item Cost Qty. Cost Qty. Cost Qty. Cost Qty. Cost
Bridge board bare PCB 25.80 1 25.80 1 25.80 1 25.80 1 25.80
Bridge board BOM 47.94 1 47.94 1 47.94 1 47.94 1 47.94
Camera 289.00 3 867.00 8 2,312.00 9 2,601.00 12 3,468.00
Lens holder 16.80 3 50.40 8 134.40 9 151.20 12 201.60
Lens 6.00 3 18.00 8 24.00 9 27.00 12 72.00
Optical filter 56.00 3 168.00 8 448.00 9 504.00 12 672.00
ODROID 46.00 1 46.00 1 46.00 3 138.00 2 92.00
Miscellaneous (*) 10.00 3 30.00 8 80.00 9 90.00 12 80.00
TOTAL 1,253.14 3,118.14 3,584.94 4,659.34
(*) fasteners, standoffs, connectors, cables, etc.
Table 5.7: Costs per set/unit based on order quantity for Bridge Board bare PCB and BOM.
Item Cost per set/unit
@ Qty = 1
Cost per set/unit
@ Qty = 10
Cost per set/unit
@ Qty = 100
Bridge board bare PCB 25.80 25.80 15.20
Bridge board BOM 47.94 39.71 31.35
The costs presented in the tables above does not account for full PCB assembly (PCBA). Nonethe-less, compared to the prices of commercial cameras presented in table 5.5, a 3-band multispectralimaging system can be built using the approach and design presented in this thesis for a total coststhat is less that half the price of the cheapest camera (2,995.00).
86
5.7 Focusing
A fundamental setting of this flexible imaging system is the possibility of changing lenses andfocusing a different distances. But focusing at a certain distance can be a challenge when workingwith multiple cameras. Since each camera has to be manually focused individually, it is important toachieve the best possible focus for all the cameras. A custom software was developed for providinga live stream of the camera capture and quantify focus, so as to provide a metric that could be usedfor achieving the best possible focus for a given target.
Pertuz et al. [98] presented an analysis and comparison of focus measure operators. They concludedthat the Laplacian-based operators presented the best overall performance for normal imaging condi-tions (i.e., without addition of noise, contrast reduction or image saturation). While all the operatorspresent different strengths and weaknesses, and they can perform differently depending on the ap-plication, the Laplacian-based operators provide a good balance between speed and robustness.
The variance of Laplacian operator [99] was chosen for its mathematical simplicity, fast computa-tional speed, and it gave consistent and precise results during testing with the image stream from thecameras. The variance of Laplacian is based on the use of a second derivative operator that allowsthe passing of high spatial frequencies which are associated with sharp edges. The second derivativeoperator used in this method is the Laplacian operator, which can be approximated using the mask
L =
0 1 01 −4 10 1 0
(5.6)
The above is a positive Laplacian operator mask in which center element of the mask should benegative and corner elements of mask should be zero. This mask is used to take out outward edges inan image. The convolution of a grayscale image with this mask highlights gray level discontinuitiesin an image. The result of this operation is an image with grayish edge lines and other discontinuitieson a dark background. Figure 5.11 shows a comparison of a sample image before and after applyingthe positive Laplacian operator.
87
Figure 5.11: Example of a sample gray scale image (left) after applying the positive Laplacian operator (right).
The focus metric calculates the variance of the absolute values, which leads to the focus measuregiven by
LAP_VAR(I) =M
∑m
N
∑n(| ∆I(m,n) | −∆I)2 (5.7)
where ∆I(m,n) is the convolution of the input image I(m,n) with the mask L and ∆I is the mean ofabsolute values given by
∆I =1
NM
M
∑m
N
∑n| ∆I(m,n) | (5.8)
Figure 5.12 shows a graph of the focus measure of a camera as the focus distance is adjusted bybringing the lens closer, going from out of focus (too far) to out of focus (too close), passing throughthe point of best focus (in-focus). Three points are highlighted in the graph: two random points inthe out of focus regions (too far and too close), and the in-focus point which is the maxima in thegraph. Figure 5.13 shows the three reference images that correspond to the highlighted points in thegraph.
88
30 40 50 60 70 80 90 100 110 120 130 140 150Focus Distance Sweep (Image Number)
55
60
65
70
75
80
85
90
Focu
s M
easu
re
Out of focus(too far)
In-focus
Out of focus(too close)
Figure 5.12: Graph of the focus measure calculated using the above algorithm as the focus is adjusted from outof focus (too far) to out of focus (too close), passing through the in-focus point. Any point outside the in-focuspoint is out of focus. The graph highlights two random out of focus points for reference, one too far and onetoo close.
(a) (b) (c)
Figure 5.13: Sample images of the three highlighted points in the graph of Figure 5.12 . (a) Out of focus (toofar). (b) In-focus. (c) Out of focus (too close).
A test was also performed to find the standard deviation of the focus measurement with the lenswith and without external vibrations. More than a 100 frames were captured for each case, and thefocus metric was calculated. When leaving the lens fixed (no vibrations) a standard deviation for thefocus measurement was 3.0302. For the lens under vibrations (lens held in place with the hand) thestandard deviation for the focus measurement was 14.2966.
In practice, each camera was focused using the same target scene and the real-time focus measure-ment software. The reported real-time focus measure was used as an indicator to adjust the lensesto the best possible focus. An acceptable threshold for this scale was set at the maximum value(in-focus) minus 5. This means, for example, that in the case of the graph shown in Figure 5.12where the maximum value (in-focus) is 85.63, a value above 80.63 will still produce an acceptablefocused image.
89
5.8 Image Alignment
Alignment of the images for this system is more complex since cameras are organized in a 3x4array, which means 2D translation is necessary. For this purpose, a subpixel registration algorithmbased on intensity [100] was used. The algorithm minimizes the mean square intensity differencebetween a reference image and a test image. The transformation of the test image, in order to achievealignment to the reference image, can be done by rigid-body motion (rotation and translation) only,or combined with isometric scaling.
The software platform ImageJ3, together with the image registration plugin StackReg4 were used foraligning the image sets. ImageJ [101] is a public Java-based image processing program developedat the National Institutes of Health. ImageJ offers multiple distributions with different features.The Fiji [102] distribution is the recommended updated general purpose option to this date, andit was the one used in this project. StackReg is a software implementation of the aforementionedalgorithm [100], written as a java plugin for ImageJ by the Biomedical Imaging Group of the ÉcolePolytechnique Fédérale de Lausanne.
The algorithm uses each image of a stack (all the images or channel that belong to the same instant)as the template for aligning the next image. The first image of the stack is used as the anchor tobegin the alignment process. The algorithm was implemented using rigid-body motion (rotation andtranslation) combined with isometric scaling. Scaling is used to account for any zooming effect thatresults from differences in focus. Performing these transformations in the images requires access todata outside the image frame. As such data is not available, it is replaced by zeroes in the outputimage. Figure 5.14 shows an example of two images being aligned. Transformations in the testimage result in edges and empty spaces filled with zeroes, which translates to black borders framingthe output image.
3https://imagej.net4http://www.epfl.ch/thevenaz/stackreg/
90
(a) (b)
Figure 5.14: Result of aligning two images. (a) Template or reference image. (b) Test image aligned to templateimage. It is possible to see the black borders filling the empty spaces after alignment.
ImageJ allows for the creation of scripts that help automate operations. A script was written in Javato run in ImageJ for taking a folder with multiples stacks of multiple images each and run a loop foraligning all the stacks. The program creates a folder where it saves all the aligned images using thesame names. Figure 5.15 shows a comparison of a composite of 3 images superposed before andafter alignment.
(a) (b)
Figure 5.15: Result of image alignment algorithm. (a) Superposition before alignment. (b) Superposition afteralignment.
A second version of the script was created where it is possible to export the transformation matricesof one stack and use it to align the rest of the stacks in the same folder. This implementationcalculates the transformation matrix once and reuses it for the rest of the stacks, thus reducingprocessing time.
91
Chapter 6
Conclusions and Future Work
The work presented in this thesis introduces the development process of a UAV-compatible multi-spectral imaging system, detailing the design, construction, testing, and validation steps. The workcovered the development of a customized fixed wing UAV, and later focuses on the development ofthe imaging system in all its revisions.
The UAV platform that was being used by the University of Alberta Aerial Robotics Group (UAARG)served as the starting point. The airframe was redesigned so as to improve construction, installation,maintenance, and more importantly, repeatability and reliability. The onboard autonomous naviga-tion system hardware was redesigned to integrate all necessary components in a compact package,and a novel redundant power supply board was added. This UAV system board solution was de-signed to be fully compatible with the Paparazzi platform, and it can be used with any fixed-wingor rotary-wing airframe. The integration on all necessary components saves time during assembly,debugging, and deployment. Additionally, the system board gives access to expansion ports for fur-ther development. Overall, better wiring management inside the plane, more efficient use of internalvolume, and the adoption of modularity, allowed the new version of the fixed wing UAV (Falcon I)to be a robust and reliable platform for research and development.
Imaging testing started with the evaluation of a small USB 2.0 CMOS camera module, with fea-tures that are similar to those found in webcams. USB 2.0 was chosen for its popularity and hard-ware/software relative simplicity compared to other interfaces such as USB 3.0, Firewire or Gigabit,which are also faster, but would increase weight and cost. The camera module was successfullystudied and tested throughout different configurations. USB camera modules are relatively sim-ple imaging systems based on an image sensor, a microcontroller, and a USB 2.0 interface. Witha relatively wide variety of image sensor sizes, which in turn provide several resolution options,USB camera modules are a compact, simple, and efficient solution for aerial imaging. These typesof cameras are an alternative option to higher end consumer cameras, offering a simpler hardware
92
that incorporates only the necessary components. This approach provides higher flexibility when itcomes to developing custom imaging systems.
With the successful results of the USB camera module testing a custom 3-band multispectral imagingsystem was developed based on similar USB camera modules and optical filters. This proof ofconcept system used 3 monochromatic camera modules with a 1.2 megapixel CMOS image sensor.The selection of this resolution was based on a basic market research that looked at commercialUAV targeted multispectral cameras. Spectral bands used in these cameras was also studied, and aselection of three bands was made for this custom system. While filters can be easily replaceable, thegreen, red, and near infrared (NIR) bands were selected. A key factor is the selection of the cameraswas the shutter. The cameras used in this system had a rolling shutter, in contrast with the globalshutter that is used in commercial multispectral cameras. It was proved that using a rolling shutterfor a 1.2 megapixels camera in standard fixed wing UAV flying conditions did not present negativeeffects. Common imaging problems associated with rolling shutter like skew and motion blur werenot visible in any of the images during normal flight. Another key factor was the use of three freerunning cameras that were capturing images at 30 frames per second each with no synchronizationbetween them. Custom software was written utilizing third party tools so as to interface with thecameras for configuration and image extraction. The developed software was optimized for low CPUusage and multiple camera interface, allowing software synchronization between the three cameras.
After successfully flying the 3-band multispectral imaging system, images were aligned using a cus-tom alignment software written in MATLAB and based on matrix correlation. The software is asimplified implementation of one-dimensional translation that looks for a correlation maxima be-tween two images. While simpler than more advanced image registration and alignment algorithms,it proved effective and fast. After alignment, the Normalized Difference Vegetation Index (NDVI)was calculated using the red and NIR images. Results showed successful and accurate detection ofvegetation with clear identification between different levels of vegetation stress.
The subsequent evolution of the project was to develop a 12-band multispectral imaging system.In this case, the same approach of multiple USB cameras was maintained. Monochromatic cameramodules with 1.2 megapixel CMOS image sensors were also used. The cameras selected for thissystem have a global shutter and an external trigger. The external trigger is a key feature of the sys-tem, allowing cameras to be accurately synchronized. A custom base board was designed containinga configurable USB HUB network and an onboard microcontroller that takes care of triggering thecameras. The system was built and tested with different configurations. The maximum number ofsimultaneous cameras for a certain configuration was determined. A way of estimating the maxi-mum number of cameras based on resolution, frame rate, and the actual practical throughput of USB2.0 (35-40 Mbytes/second) was provided. This customizable multi-camera imaging system providesa flexible and versatile development platform for multispectral imaging applications. A method forachieving in-focus for all cameras was presented, based on a custom software that quantifies focus inreal-time. Using this focus metric it was possible to bring all cameras to the best possible focus us-
93
ing a reference target scene. Finally, an alignment method was also introduced based on the ImageJimage processing software. The alignment algorithm uses subpixel registration based on intensity,and can operate using two-dimensional translation, rotation, and isometric scaling.
In terms of costs, the 12-band multispectral imaging hardware presented in this work can achievea considerably lower cost when compared to similar UAV compatible commercial multispectralcameras for the same number of bands. The custom bridge board is a multifunctional and importantpiece of hardware with a relatively low cost, which obviously drops as order quantities increase. Alower overall cost, higher development flexibility and control, and capacity for expansion, make thissystem a viable alternative to commercial off-the-shelf systems.
The whole process presented in this work followed a careful methodology that started with the test-ing of a single USB 2.0 camera module. After validation, this type of camera became the core of amulti-camera imaging system. The first revision or proof of concept consisted on a combination of3 cameras and optical filters for the development a 3-band multispectral imaging system. After suc-cessful testing, the designed evolved into a 12-band multispectral imaging system with integratedUSB HUB and automatic camera triggering system. The work done included hardware and soft-ware development, and provides a full imaging solution for custom applications. All the softwaredevelopment and code used in this project and referenced in this thesis can be found in the followinggithub repository: https://github.com/JLMARIN/imaging.git.
6.1 Future Work
Future work of this thesis would contemplate the integration of small single board computers (SBC)into the 12-band multispectral imaging system. With the considerably low cost of SBCs, it shouldbe possible to find a small but powerful board with minimal components, that would connect viaUSB to a group of cameras through the camera bridge board. The SCB would configure and extractimages from the cameras, which are all synchronized thanks to the trigger system. With a networkof several SBCs connected via a high bandwidth channel like ethernet or WiFi, a master SBC couldgather all images from the other computers and have them stored in an SD card or transferred to anexternal system. The imaging platform and processing computer can be packaged together into acompact device.
A further hardware development step along the multi-camera imaging system would be to design aPCB with image sensors interfaced directly with an FPGA. This approach would allow costs to dropconsiderably, and would remove the relatively slow and limited USB 2.0 interface. With the FPGAtalking directly to the sensors it would be possible to retrieve images from all of them simultaneouslyand store the images in an SD card.
Finally, it would be interesting and useful to make the system capable of integration with nearinfrared (NIR) and short-wave infrared (SWIR) cameras, expanding the spectrum coverage of the
94
overall system. If a trigger is present in such cameras, it can be synchronized with the trigger systemthat controls the Bridge board, and have the interface computer retrieve and save the images.
95
Bibliography
[1] J. Wilford and J. Creasey, “Landsat thematic mapper,” Geophysical and Remote Sensing
Methods for Regolith Exploration, pp. 6–12, 2002.
[2] B. L. Markham and J. L. Barker, “Spectral characterization of the landsat thematic mappersensors,” International Journal of Remote Sensing, vol. 6, no. 5, pp. 697–716, 1985.
[3] J. Vogelmann, D. Helder, R. Morfitt, M. Choate, and J. Merchant, “Characterization ofLandsat Thematic Mapper radiometry and geometry for land cover analysis,” IGARSS ’98.
Sensing and Managing the Environment. 1998 IEEE International Geoscience and Remote
Sensing. Symposium Proceedings. (Cat. No.98CH36174), vol. 5, pp. 2405–2407 vol.5,1998. [Online]. Available: http://ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=702228
[4] D. L. Williams, J. R. Irons, B. L. Markham, R. F. Nelson, D. L. Toll, R. S. Latty, and M. L.Stauffer, “A statistical evaluation of the advantages of Landsat Thematic Mapper data incomparison to Multispectral Scanner data,” IEEE Transactions on Geoscience and Remote
Sensing, vol. GE-22, no. 3, pp. 294–302, 1984.
[5] R. D. Lees, W. R. Lettis, and R. Bernstein, “Evaluation of Landsat Thematic Mapper Imageryfor Geologic Applications,” Proceedings of the IEEE, vol. 73, no. 6, pp. 1108–1117, 1985.
[6] D. Lamb, “The use of qualitative airborne multispectral imaging for managing agriculturalcrops-a case study in south-eastern Australia,” Australian Journal of Experimental Agricul-
ture, vol. 40, no. 5, pp. 725–738, 2000.
[7] Y. Huang, S. J. Thomson, Y. Lan, and S. J. Maas, “Multispectral imaging systems for air-borne remote sensing to support agricultural production management,” International Journal
of Agricultural and Biological Engineering, vol. 3, no. 1, pp. 50–62, 2010.
[8] S. Nebiker, N. Lack, M. Abächerli, and S. Läderach, “Light-weight multispectral uav sen-sors and their capabilities for predicting grain yield and detecting plant diseases,” Interna-
tional Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences,vol. 2016-Janua, no. July, pp. 963–970, 2016.
96
[9] P. P. J. Roosjen, J. M. Suomalainen, H. M. Bartholomeus, and J. G. P. W. Clevers, “Hyperspec-tral reflectance anisotropy measurements using a pushbroom spectrometer on an unmannedaerial vehicle-results for barley, winter wheat, and potato,” Remote Sensing, vol. 8, no. 10,pp. 1–16, 2016.
[10] S. Candiago, F. Remondino, M. D. Giglio, M. Dubbini, and M. Gattelli, “Evaluating Mul-tispectral Images and Vegetation Indices for Precision Farming Applications from UAV Im-ages,” Remote Sensing, vol. 7, no. Vi, pp. 4026–4047, 2015.
[11] K. Themistocleous, A. Agapiou, B. Cuca, and D. G. Hadjimitsis, “UnmannedAerial Systems and Spectroscopy for Remote Sensing Applications in Archaeology,”ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial
Information Sciences, vol. XL-7/W3, no. May, pp. 1419–1423, 2015. [Online]. Available:http://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XL-7-W3/1419/2015/
[12] D. Turner, A. Lucieer, Z. Malenovský, D. H. King, and S. A. Robinson, “Spatial co-registration of ultra-high resolution visible, multispectral and thermal images acquired with amicro-UAV over antarctic moss beds,” Remote Sensing, vol. 6, no. 5, pp. 4003–4024, 2014.
[13] L. Hassan-Esfahani, A. Torres-Rua, A. Jensen, and M. McKee, “Assessment of surface soilmoisture using high-resolution multi-spectral imagery and artificial neural networks,” Remote
Sensing, vol. 7, no. 3, pp. 2627–2646, 2015.
[14] Sentek Systems, “GEMS Multispectral Sensor,” 2017. [Online]. Available: http://precisionaguavs.com/sensors/. [Accessed: 2017-06-23].
[15] Sentera LLC, “Sentera Quad Sensor,” 2017. [Online]. Available: https://sentera.com/product/sentera-quad-sensor/. [Accessed: 2017-06-23].
[16] Parrot SA, “Parrot Sequoia,” 2017. [Online]. Available: https://www.parrot.com/ca/business-solutions/parrot-sequoia#parrot-sequoia-. [Accessed: 2017-06-22].
[17] B. E. A. Saleh and M. C. Teich, Fundamentals of Photonics, 2nd ed. Hoboken, N.J.: Wiley-Interscience, 2007.
[18] Wikimedia Commons, “Angle of view,” 2012. [Online]. Available: en.wikipedia.org/wiki/File:Angle_of_view.svg. [Accessed: 2017-06-11].
[19] Panasonic, “How Focal Length Affects Viewing Angle.” [Online]. Available: http://av.jpn.support.panasonic.com/support/global/cs/dsc/knowhow/knowhow12.html. [Accessed:2017-06-11].
[20] LEMKE, “Do you know the difference between the CCD and CMOS chip technology?”2014. [Online]. Available: http://www.lemkevision.com/cn/news/. [Accessed: 2017-06-11].
97
[21] FLIR Knowledge Base, “Key differences between rolling shutter and frame (global) shutter,”2015. [Online]. Available: www.ptgrey.com/kb/10028#. [Accessed: 2017-11-06].
[22] T. York and R. Jain, “Fundamentals of Image Sensor Performance,” pp. 1–8, 2011. [Online].Available: http://www.cse.wustl.edu/~jain/cse567-11/ftp/imgsens/index.html
[23] P. Swain and D. Cheskis, “Back-Illuminated Image Sensors Come to the Forefront,” 2008.[Online]. Available: www.photonics.com/Article.aspx?AID=34685
[24] V. Suntharalingam, R. Berger, S. Clark, J. Knecht, A. Messier, K. Newcomb, D. Rathman,R. Slattery, A. Soares, C. Stevenson, K. Warner, D. Young, P. Ang, Lin, B. Mansoorian, andD. Shaver, “A 4-side tileable back illuminated 3d-integrated Mpixel CMOS image sensor,”Digest of Technical Papers - IEEE International Solid-State Circuits Conference, pp. 38–40,2009.
[25] Edmund Optics, “Introduction to Modulation Transfer Function.”[Online]. Available: www.edmundoptics.com/resources/application-notes/optics/introduction-to-modulation-transfer-function/. [Accessed: 2017-11-16].
[26] A. A. Michelson, Studies in Optics. Chicago, Ill: The University of Chicago Press, 1927.
[27] Dxomark, “MTF.” [Online]. Available: https://www.dxomark.com/glossary/mtf-2/. [Ac-cessed: 2017-11-27].
[28] ISO, “ISO 12233:2017, Photography – Electronic still picture imaging – Resolution and spa-tial frequency responses,” 2017.
[29] P. D. Burns, “Slanted-Edge MTF for Digital Camera and Scanner Analysis,” in Proc. IS&T
2000 PICS Conference, vol. 3, 2000, pp. 135–138.
[30] S. E. Reichenbach, S. K. Park, and R. Narayanswamy, “Characterizing digital image acquisi-tion devices,” Optical Engineering, vol. 30, no. 2, pp. 170–177, 1991.
[31] D. Williams, “Benchmarking of the ISO 12233 Slanted-edge Spatial Frequency ResponsePlug-in Don,” in IS&T’s 1998 PICS Conference, 1998.
[32] P. D. Burns, “Refined slanted-edge measurement for practical camera and scanner testing,”IS&T PICS Conference, pp. 191–195, 2002.
[33] P. D. Burns and D. Williams, “Digital Camera and Scanner Performance Evaluation : Stan-dards and Software,” in SPIE/IS&T Electronic Imaging Symposium, 2014, pp. 1–41.
[34] P. D. Burns and D. Williams, “Sampling Efficiency in Digital Camera Performance Stan-dards,” in Proc. SPIE, vol. 6808, 2008, pp. 1 – 5.
98
[35] M. H. Schimek, B. Dirks, H. Verkuil, and M. Rubli, Video for Linux two API specification -
Revision 0.24, 2008.
[36] L. Yinli, Y. Hongli, and Z. Pengpeng, “The Implementation of Embedded Image Acquisi-tion Based on V4L2,” 2011 International Conference on Electronics, Communications and
Control (Icecc), pp. 549–552, 2011.
[37] S. D. Burks and J. M. Doe, “GStreamer as a framework for image processing applicationsin image fusion,” Proc. SPIE 8064, Multisensor, Multisource Information Fusion:
Architectures, Algorithms, and Applications 2011, vol. 8064, pp. 80 640M–80 640M–7,2011. [Online]. Available: http://dx.doi.org/10.1117/12.884797
[38] Gstreamer.freedesktop.org, “What is GStreamer?” [Online]. Available: https://gstreamer.freedesktop.org/documentation/application-development/introduction/gstreamer.html. [Ac-cessed: 2018-01-18].
[39] Tetracam Inc., “Tetracam ADC Micro,” 2017. [Online]. Available: http://tetracam.com/Products-ADC_Micro.htm. [Accessed: 2017-06-22].
[40] Tetracam Inc., “Tetracam Mini MCA,” 2017. [Online]. Available: http://www.tetracam.com/Products-Mini_MCA.htm. [Accessed: 2017-06-22].
[41] A. Burkart, S. Cogliati, A. Schickling, and U. Rascher, “A novel UAV-Based ultra-lightweight spectrometer for field spectroscopy,” IEEE Sensors Journal, vol. 14, no. 1, pp. 62–67, 2014.
[42] J. Suomalainen, N. Anders, S. Iqbal, G. Roerink, J. Franke, P. Wenting, D. Hünniger,H. Bartholomeus, R. Becker, and L. Kooistra, “A lightweight hyperspectral mapping sys-tem and photogrammetric processing chain for unmanned aerial vehicles,” Remote Sensing,vol. 6, no. 11, pp. 11 013–11 030, 2014.
[43] F. Agüera, F. Carvajal, and M. Pérez, “Measuring Sunflower Nitrogen Status From an Un-manned Aerial Vehicle-Based System and an on the Ground Device,” ISPRS - International
Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol.XXXVIII-1/, no. September, pp. 33–37, 2011.
[44] J. S. West, C. Bravo, R. Oberti, D. Lemaire, D. Moshou, and H. A. McCartney, “ThePotential of Optical Canopy Measurement for Targeted Control of Field Crop Diseases,”Annual Review of Phytopathology, vol. 41, no. 1, pp. 593–614, 2003. [Online]. Available:http://www.annualreviews.org/doi/10.1146/annurev.phyto.41.121702.103726
[45] S. Jacquemoud and S. L. Ustin, “Leaf optical properties: a state of the art,” Proc. 8th
International Symposium Physical Measurements & Signatures in Remote Sensing, no.January, pp. 223–232, 2001. [Online]. Available: citeulike-article-id:7215646
99
[46] Y. Wang, G. Li, J. Ding, Z. Guo, S. Tang, R. Liu, and J. Chen, “A combinedGLAS and MODIS estimation of the global distribution of mean forest canopy height,”Remote Sensing of Environment, vol. 174, pp. 24–43, 2016. [Online]. Available:http://dx.doi.org/10.1016/j.rse.2015.12.005
[47] Y. Knyazikhin, M. A. Schull, P. Stenberg, M. Mõttus, M. Rautiainen, Y. Yang, A. Marshak,P. Latorre Carmona, R. K. Kaufmann, P. Lewis, M. I. Disney, V. Vanderbilt, A. B.Davis, F. Baret, S. Jacquemoud, A. Lyapustin, and R. B. Myneni, “Hyperspectral remotesensing of foliar nitrogen content.” Proceedings of the National Academy of Sciences of
the United States of America, vol. 110, no. 3, pp. E185–92, 2013. [Online]. Available:http://www.pnas.org/content/110/3/E185.short
[48] M. Kneubühler, B. Koetz, S. Huber, M. E. Schaepman, and N. E. Zimmermann, “Space-based spectrodirectional measurements for the improved estimation of ecosystem variables,”Canadian Journal of Remote Sensing, vol. 34, no. 3, pp. 192–205, 2008.
[49] Z. Wang, C. A. Coburn, X. Ren, and P. M. Teillet, “Effect of soil surface roughness andscene components on soil surface bidirectional reflectance factor,” Canadian Journal of Soil
Science, vol. 92, no. 2, pp. 297–313, 2012.
[50] P. P. J. Roosjen, H. M. Bartholomeus, and J. G. P. W. Clevers, “Effects of soil moisturecontent on reflectance anisotropy - Laboratory goniometer measurements and RPV modelinversions,” Remote Sensing of Environment, vol. 170, pp. 229–238, 2015. [Online].Available: http://dx.doi.org/10.1016/j.rse.2015.09.022
[51] A. Bannari, D. Morin, F. Bonn, and A. R. Huete, “A review of vegetation indices,”Remote Sensing Reviews, vol. 13, no. 1, pp. 95–120, 1995. [Online]. Available:http://dx.doi.org/10.1080/02757259509532298
[52] A. Matese, F. Capraro, J. Primicerio, G. Gualato, S. Gennaro, and G. Agati, “Mapping ofvine vigor by UAV and anthocyanin content by a non- destructive fluorescence technique,” inStafford J.V. (eds) Precision agriculture ’13. Wageningen: Wageningen Academic Publish-ers, 2013, pp. 201–208.
[53] A. a. Gitelson, Y. Gritz, and M. N. Merzlyak, “Relationships between leaf chlorophyll contentand spectral reflectance and algorithms for non-destructive chlorophyll assessment in higherplant leaves.” Journal of plant physiology, vol. 160, no. 3, pp. 271–82, 2003.
[54] H. Zheng, X. Zhou, T. Cheng, X. Yao, Y. Tian, W. Cao, and Y. Zhu, “Evaluation of a UAV-based Hyperspectral Frame Camera for Monitoring the Leaf Nitrogen Concentration in Rice,”in Geoscience and Remote Sensing Symposium (IGARSS), 2016, pp. 7350–7353.
100
[55] I. Pölönen, H. Saari, J. Kaivosoja, E. Honkavaara, and L. Pesonen, “Hyperspectral imagingbased biomass and nitrogen content estimations from light-weight UAV,” in Proceedings of
SPIE - The International Society for Optical Engineering, 2013.
[56] V. P. Polischuk, T. M. Shadchina, T. I. Kompanetz, I. G. Budzanivskaya, A. L.Boyko, and A. A. Sozinov, “Changes in reflectance spectrum characteristic of nicotianadebneyi plant under the influence of viral infection,” Archives Of Phytopathology
And Plant Protection, vol. 31, no. 1, pp. 115–119, 1997. [Online]. Available:http://dx.doi.org/10.1080/03235409709383221
[57] R. Näsi, E. Honkavaara, P. Lyytikäinen-Saarenmaa, M. Blomqvist, P. Litkey, T. Hakala,N. Viljanen, T. Kantola, T. Tanhuanpää, and M. Holopainen, “Using UAV-based photogram-metry and hyperspectral imaging for mapping bark beetle damage at tree-level,” in Remote
Sensing, no. 11, 2013, pp. 15 467–15 493.
[58] F. Garcia-Ruiz, S. Sankaran, J. M. Maja, W. S. Lee, J. Rasmussen, and R. Ehsani,“Comparison of two aerial imaging platforms for identification of Huanglongbing-infectedcitrus trees,” Computers and Electronics in Agriculture, vol. 91, pp. 106–115, 2013. [Online].Available: http://dx.doi.org/10.1016/j.compag.2012.12.002
[59] J. Flexas, J. M. Escalona, and H. Medrano, “Water stress induces different levels of photo-synthesis and electron transport rate regulation in grapevines,” Plant, Cell and Environment,vol. 22, no. 1, pp. 39–48, 1999.
[60] J. Flexas, J. M. Escalona, S. Evain, J. Gulías, I. Moya, C. B. Osmond, and H. Medrano,“Steady-state chlorophyll fluorescence (Fs) measurements as a tool to follow variations ofnet CO2 assimilation and stomatal conductance during water-stress in C3 plants,” European
Space Agency, (Special Publication) ESA SP, no. 527, pp. 26–29, 2002.
[61] P. J. Zarco-Tejada, V. González-Dugo, and J. A. J. Berni, “Fluorescence, temperature andnarrow-band indices acquired from a UAV platform for water stress detection using amicro-hyperspectral imager and a thermal camera,” Remote Sensing of Environment, vol.117, pp. 322–337, 2012. [Online]. Available: http://dx.doi.org/10.1016/j.rse.2011.10.007
[62] E. Honkavaara, M. A. Eskelinen, I. Polonen, H. Saari, H. Ojanen, R. Mannila, C. Holmlund,T. Hakala, P. Litkey, T. Rosnell, N. Viljanen, and M. Pulkkanen, “Remote Sensing of 3-D Geometry and Surface Moisture of a Peat Production Area Using Hyperspectral FrameCameras in Visible to Short-Wave Infrared Spectral Ranges Onboard a Small UnmannedAirborne Vehicle (UAV),” IEEE Transactions on Geoscience and Remote Sensing, vol. 54,no. 9, pp. 5440–5454, 2016.
[63] R. Bryant, D. Thoma, S. Moran, C. Holifield, D. Goodrich, T. Keefer, G. Paige, David,Williams, and S. Skirvin, “Evaluation of Hyperspectral, Infrared Temperature and Radar
101
Measurements for Monitoring Surface Soil Moisture,” First Interagency Conference in the
Watersheds, pp. 528–533, 2003.
[64] M. P. Finn, M. D. Lewis, D. D. Bosch, M. Giraldo, K. Yamamoto, D. G. Sullivan, R. Kincaid,R. Luna, G. K. Allam, C. Kvien, and M. S. Williams, “Remote Sensing of Soil MoistureUsing Airborne Hyperspectral Data,” GIScience & Remote Sensing, vol. 48, no. 4, pp. 522–540, 2011.
[65] C. S. Allen, M. P. S. Krekeler, C. Scott Allen, and M. P. S. Krekeler, “Reflectance spectra ofcrude oils and refined petroleum products on a variety of common substrates,” Proceedings
of SPIE, vol. 7687, no. May 2012, pp. 76 870L–1–76 870L–13, 2010.
[66] B. Hörig, F. Kühn, F. Oschütz, and F. Lehmann, “HyMap hyperspectral remote sensing todetect hydrocarbons,” International Journal of Remote Sensing, vol. 22, no. 8, pp. 1413–1422, 2001.
[67] I. Leifer, W. J. Lehr, D. Simecek-Beatty, E. Bradley, R. Clark, P. Dennison, Y. Hu,S. Matheson, C. E. Jones, B. Holt, M. Reif, D. A. Roberts, J. Svejkovsky, G. Swayze,and J. Wozencraft, “State of the art satellite and airborne marine oil spill remote sensing:Application to the BP Deepwater Horizon oil spill,” Remote Sensing of Environment, vol.124, pp. 185–209, 2012. [Online]. Available: http://dx.doi.org/10.1016/j.rse.2012.03.024
[68] R. N. Clark, G. A. Swayze, I. Leifer, K. E. Livo, R. Kokaly, T. Hoefen, S. Lundeen, M. East-wood, R. O. Green, N. Pearson, C. Sarture, I. Mccubbin, D. Roberts, E. Bradley, D. Steele,T. Ryan, R. Dominguez, and Airborne Visible/Infrared Imaging Spectrometer (AVIRIS)Team, “A method for quantitative mapping of thick oil spills using imaging spectroscopy,”Tech. Rep. 1167, 2010.
[69] B. Rivard, J. Feng, V. Bushan, and M. Lipsett, “Infrared reflectance hyperspectral features ofAthabasca oil sand ore and froth,” Workshop on Hyperspectral Image and Signal Processing,
Evolution in Remote Sensing, pp. 0–3, 2011.
[70] B. Rivard, D. Lyder, J. Feng, A. Gallie, E. Cloutis, P. Dougan, S. Gonzalez, D. Cox, andM. G. Lipsett, “Bitumen content estimation of Athabasca oil sand from broad band infraredreflectance spectra,” Canadian Journal of Chemical Engineering, vol. 88, no. 5, pp. 830–838,2010.
[71] D. Hausamann, W. Zirnig, G. Schreier, and P. Strobl, “Monitoring of gas pipelines- a civil UAV application,” Aircraft Engineering and Aerospace Technology, vol. 77,no. 5, pp. 352–360, 2005. [Online]. Available: http://www.emeraldinsight.com/doi/10.1108/00022660510617077
102
[72] J. Tuominen and T. Lipping, “Detection of Environmental Change Using Hyperspectral Re-mote Sensing at Olkiluoto Repository Site,” POSIVA OY, Eurajoki, Finland, Tech. Rep.,2011.
[73] A. J. S. McGonigle, A. Aiuppa, G. Giudice, G. Tamburello, A. J. Hodson, and S. Gurrieri,“Unmanned aerial vehicle measurements of volcanic carbon dioxide fluxes,” Geophysical
Research Letters, vol. 35, no. 6, pp. 3–6, 2008.
[74] A. Agapiou, D. G. Hadjimitsis, and D. D. Alexakis, “Evaluation of broadband and narrow-band vegetation indices for the identification of archaeological crop marks,” Remote Sensing,vol. 4, no. 12, pp. 3892–3919, 2012.
[75] M. Doneus, G. Verhoeven, C. Atzberger, M. Wess, and M. Rus, “New ways to extract archae-ological information from hyperspectral pixels,” Journal of Archaeological Science, vol. 52,pp. 84–96, 2014.
[76] D. C. Cowley, C. Moriarty, G. Geddes, G. L. Brown, T. Wade, and C. J. Nichol, “UAVsin Context: Archaeological Airborne Recording in a National Body of Survey and Record,”Drones, vol. 2, no. 2, pp. 1–16, 2018.
[77] C. Moriarty, “Deploying Multispectral Remote Sensing for Multitemporal Analysis of Ar-chaeological Crop Stress at Ravenshall, Fife,” Master’s Thesis, University of Edinburgh,2017.
[78] H. Chao, Y. Cao, and Y. Chen, “Autopilots for small unmanned aerial vehicles: A survey,”International Journal of Control, Automation and Systems, vol. 8, no. 1, pp. 36–44, 2010.
[79] H. Lim, J. Park, D. Lee, and H. J. Kim, “Build your own quadrotor: Open-source projectson unmanned aerial vehicles,” IEEE Robotics and Automation Magazine, vol. 19, no. 3, pp.33–45, 2012.
[80] E. Ebeid, M. Skriver, and J. Jin, “A Survey on Open-Source Flight Control Platforms ofUnmanned Aerial Vehicle,” in Euromicro Symposium on Digital Systems Design. IEEE,2017.
[81] B. Remes, P. Esden-Tempski, F. Van Tienen, E. Smeur, C. De Wagter, and G. De Croon,“Lisa-S 2.8g autopilot for GPS-based flight of MAVs,” IMAV 2014, International Micro Air
Vehicle Conference and Competition 2014, pp. 280–285, 2014.
[82] R. W. Beard, D. Kingston, M. Quigley, D. Snyder, R. Christiansen, W. Johnson, T. McLain,and M. Goodrich, “Autonomous Vehicle Technologies for Small Fixed-Wing UAVs,” Journal
of Aerospace Computing, Information, and Communication, vol. 2, no. 1, pp. 92–108, 2005.
103
[83] D. Stojcsics and A. Molnár, “AirGuardian - UAV hardware and software system for smallsize UAVs,” International Journal of Advanced Robotic Systems, vol. 9, pp. 1–8, 2012.
[84] MicroPilot Company, “MP2128-3X MicroPilot’s Triple Redundant UAV Autopilot,” Tech.Rep. [Online]. Available: https://www.micropilot.com/pdf/white-papers/mp21283x.pdf
[85] J. Marin, R. Augustine, M. Lipsett, and Duncan G. Elliott, “An Open Source ComprehensiveSystem Board for UAVs with Redundant Power,” in Unmanned Canada 2016 Conference,Edmonton, 2016.
[86] B. Gati and Paparazzi Community, “Open Source Autopilot for Academic Research - ThePaparazzi System,” Proceeding of the American Control Conference 2013, pp. 1478–1481,2013.
[87] P. Brisset, A. Drouin, M. Gorraz, P.-s. Huard, and J. Tyler, “The Paparazzi Solution,” MAV
2006, 2nd US-European Competition and Workshop on Micro Air Vehicles, pp. 1–15, 2006.
[88] G. Hattenberger, M. Bronz, and M. Gorraz, “Using the Paparazzi UAV System for ScientificResearch,” IMAV 2014, International Micro Air Vehicle Conference and Competition 2014,pp. 247–252, 2014.
[89] P. Burns, “sfrmat3: SFR evaluation for digital cameras and scanners,” 2015. [Online].Available: http://losburns.com/imaging/software/SFRedge/. [Accessed: 2017-04-10].
[90] J. P. Jhan, J. Y. Rau, N. Haala, and M. Cramer, “Investigation of parallax issues for multi-lensmultispectral camera band co-registration,” International Archives of the Photogrammetry,
Remote Sensing and Spatial Information Sciences - ISPRS Archives, vol. 42, no. 2W6, pp.157–163, 2017.
[91] A. Bhandari, A. Kumar, and G. Singh, “Feature Extraction using Normalized DifferenceVegetation Index (NDVI): A Case Study of Jabalpur City,” Procedia Technology, vol. 6, pp.612–621, 2012.
[92] QImaging, “Global Shutter vs. Rolling Shutter Readouts,” pp. 1–9, 2014. [Online]. Available:https://www.qimaging.com/ccdorscmos/pdfs/RollingvsGlobalShutter.pdf
[93] W. Hong, D. Wei, and A. U. Batur, “Video stabilization and rolling shutter distortion re-duction,” in Proceedings - International Conference on Image Processing, ICIP, 2010, pp.3501–3504.
[94] Z. W. Pan, H. L. Shen, C. Li, S. J. Chen, and J. H. Xin, “Fast Multispectral Imaging by SpatialPixel-Binning and Spectral Unmixing,” IEEE Transactions on Image Processing, vol. 25,no. 8, pp. 3612–3625, 2016.
104
[95] Arduino, “Arduino - Home Page,” 2018. [Online]. Available: https://www.arduino.cc/.[Accessed: 2018-02-05].
[96] J. C. Martínez-Santos, O. Acevedo-Patino, and S. H. Contreras-Ortiz, “Influence of Arduinoon the Development of Advanced Microcontrollers Courses,” Revista Iberoamericana de Tec-
nologias del Aprendizaje, vol. 12, no. 4, pp. 208–217, 2017.
[97] A. Orebaugh, G. Ramirez, and J. Beale, Wireshark & Ethereal Network Protocol Analyzer
Toolkit. Elsevier, 2006.
[98] S. Pertuz, D. Puig, and M. A. Garcia, “Analysis of focus measure operators forshape-from-focus,” Pattern Recognition, vol. 46, no. 5, pp. 1415–1432, 2013. [Online].Available: http://dx.doi.org/10.1016/j.patcog.2012.11.011
[99] J. Pech-Pacheco, G. Cristobal, J. Chamorro-Martinez, and J. Fernandez-Valdivia, “Diatomautofocusing in brightfield microscopy: a comparative study,” Proceedings 15th International
Conference on Pattern Recognition. ICPR-2000, vol. 3, pp. 314–317, 2000. [Online].Available: http://ieeexplore.ieee.org/document/903548/
[100] P. Thévenaz, U. E. Ruttimann, and M. Unser, “A pyramid approach to subpixel registrationbased on intensity,” IEEE Transactions on Image Processing, vol. 7, no. 1, pp. 27–41, 1998.
[101] C. T. Rueden, J. Schindelin, M. C. Hiner, B. E. DeZonia, A. E. Walter, E. T. Arena, andK. W. Eliceiri, “ImageJ2: ImageJ for the next generation of scientific image data,” BMC
Bioinformatics, vol. 18, no. 1, pp. 1–26, 2017.
[102] J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. Preibisch,C. Rueden, S. Saalfeld, B. Schmid, J. Y. Tinevez, D. J. White, V. Hartenstein, K. Eliceiri,P. Tomancak, and A. Cardona, “Fiji: An open-source platform for biological-image analysis,”Nature Methods, vol. 9, no. 7, pp. 676–682, 2012.
105
Appendix A
UAV System Board User Guide
Autopilot Embedded System Board - Falcon IUser Guide
Jorge Marin and Duncan G. Elliott
Rev. 1.5 - March 29, 2018
Introduction
This user guide introduces the Autopilot Embedded System Board as part of the Falcon I project developedin the University of Alberta, and describes the use and development capabilities for Unmanned Aerial Vehicle(UAV) applications. This project was developed with the support and assistance of Rijesh Augustine, andthe University of Alberta Aerial Robotics Group (UAARG).
106
Scope
This guide provides details on the Falcon I Autopilot system. It is made up of four main sections:
• Section 1 - describes the Falcon I Autopilot main features.
• Section 2 - provides instructions to power up the Falcon I Autopilot board.
• Section 3 - provides an overview of the Falcon I Autopilot board.
• Section 4 - describes the Falcon I Autopilot board components.
Related Items
• STMicroelectronics STM32F405 Datasheet (http://www.st.com/content/ccc/resource/technical/document/datasheet/ef/92/76/6d/bb/c2/4f/f7/DM00037051.pdf/files/DM00037051.pdf/jcr:content/translations/en.DM00037051.pdf)
107
Contents
1 Specifications 4
1.1 Electrostatic Warning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Power Supply Warning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Power Up 5
2.1 Power Up the Board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 Hardware Introduction 6
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 Equipment List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.3 Board Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4 Board Components 7
4.1 Board Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.2 Function Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2.1 Microcontroller Unit (MCU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2.2 Power Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2.3 Inertial Measurement Unit (IMU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2.4 GPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.2.5 XBee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.2.6 Data-Logger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.2.7 CAN Transceiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2.8 Motor/Servo Outputs (PWM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2.9 Connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2.10 Jumpers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.3 Board Pictures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.4 Autopilot Board Schematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.5 Autopilot Board Bill-of-Materials (BOM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5 Revision History 48
108
1 SpecificationsTable 1.1: Board Specifications.
Characteristic SpecificationsTemperature:- Operating- Storage
0ºC to +70ºC-40ºC to +85ºC
Relative Humidity 0 to 90% (non-condensing)RoHS status Compliant
1.1 Electrostatic Warning
Warning: ESD-Sensitive Electronic Equipment!
The board system must not be subject to high electrostatic potentials.
It is strongly recommended to use a grounding strap or similar ESD pro-tective device when handling the board in hostile ESD environments (forexample, places with synthetic carpets). Avoid touching the componentpins or any other metallic element on the board.
1.2 Power Supply Warning
Warning: Hardware Power Supply Limitation
Power supply must be 5VDC. Using a power adapter greater than 5VDCmay damage the board.
Warning: Hardware Power Budget
Using the USB port as the main power source (max. 500 mA) is acceptablewhen using the on-board peripherals only.
When connecting external peripherals or add-on boards, it is recommended the use of an external powersource connected to the J9 POWER header (can provide up to 1A on the 3.3V node).
109
2 Power Up
Several power source options are available to power up the Falcon I Autopilot board.
The board can be:
• USB-powered through the Micro USB connector (USB1 connector).
• Powered through an external 5V source connected via the J9 POWER header. The external powersource must be able to supply 200mA just for the board to work with the basic configuration. Additionalperipherals or boards connected might require more supply current.
• Powered through an external 5V source connected via the J5 expansion header - pins 1 and 2. Sameconditions as above.
Warning: The Falcon I Autopilot board runs at 3.3V. The maximum voltage that the I/O pins can tolerateis 3.3V. Providing higher voltages (e.g. 5V) to an I/O pin could damage the board.
2.1 Power Up the Board
Take the board and connect 5V and GND through the J9 POWER header (check pinout information for J9to see what pins to use).
Table 2.1: Electrical Characteristics.
Electrical Parameter ValuesInput Voltage 5VDCMax DC 3.3V Current Available 1AI/O Voltage 3.3V only
2.2 Programming
The Falcon I Autopilot board is fully compatible with the Lisa MX from 1 Bit Square, and therefore canbe programmed using the same external programmer. This programmer is called “Black Magic Probe”, andconnects directly to the Falcon I Autopilot board through the JTAG port. Connect the other end to a freeUSB port of your PC using a mini-USB to USB cable.
Open PaparazziUAV, and configure and program the board using the same source files used with the LisaMX.
110
3 Hardware Introduction
3.1 Introduction
The Falcon I Autopilot board is a fully-featured development platform for UAV applications. It integratesall major components needed for UAV autopilot control.
3.2 Equipment List
The Falcon I Autopilot board is built around the integration of a Cortex®-M4-based microcontroller withfloating-point unit, 3-axis accelerometer, gyroscope, and magnetometer, barometer, GPS unit, secondarymicrocontroller with SD card for data-logging, 8 PWM outputs for motors/servos and expansion headersand connectors.
3.3 Board FeaturesTable 3.1: Board Specifications.
Characteristics SpecificationsPCB characteristics 90 x 52 x 11mm (4-layers)Microcontroller STM32F405VGT6 - 168MHz 32-bit ARM® Cortex® M4 MCU
with 192KB RAM, 1024KB Flash, and floating point unit(FPU)
USB One Micro-USB10 DOF IMU 3 axis accelerometer
3 axis gyroscope3 axis magnetometerbarometer
Remote Control Two UART ports for remote control receiversTelemetry One socket for XBee radioGPS One uBlox MAX-7Q GPS unit
The GPS on-board circuitry includes:- SAW filter- Optional LNA- Selectable uFL or SMA connector for GPS antenna
Debug Port One JTAG interface connectorPWM Outputs Eight PWM outputs for motors/servosData-logger Secondary MCU and micro-SD card socket for telemetry
data-loggingExpansion Connectors - One high speed SPI interface for high speed hardware
expansion- Two I2C interfaces for actuators and sensors- One CAN interface (with onboard transceiver) for actuatorsand sensors- Three General Purpose IO pins- Three analog inputs- One UART port with adjustable level shifter- One 48-pin header with access to multiple MCU pins
Board Supply Voltage 5V from USB or J9 POWER headerOn-board power regulation to 3.3V is performed by a LDOlinear voltage regulator
User Interface Bind buttonEight indicator LEDs for status check
111
4 Board Components
4.1 Board Overview
The Falcon I Autopilot board integrates several peripherals and interface connectors, as shown in Figures4.1 and 4.2.
Figure 4.1: Falcon I Autopilot Board Overview (Top).
Figure 4.2: Falcon I Autopilot Board Overview (Bottom).
112
The Falcon I Autopilot board is equipped with the following interface connectors:
• J3: PWM outputs for motors/servos
• J5: Expansion 48-pin header with access to power, GPIO, and communication ports
• J6: UART3 port (connected to GPS)
• J7: Control switches connector
• J9: Main power supply
• J10: Expansion connector with two UART ports (UART1 and UART5) for RC receivers and one I2Cport (I2C1)
• J11: I2C 3.3V port (I2C2)
• J12: I2C 5V port (I2C2)
• J13: UART1 port (for RC receiver)
• J14: UART5 port (for RC receiver)
• J15: ISP connector for data-logger MCU
• J16: FTDI connector for data-logger MCU
• J17: uFL connector for GPS antenna
• J18: SMA connector for GPS antenna
• J19: SPI port
• J20: UART2 port (connected to XBee)
• J21: UART port with voltage level converter
• J22: Secondary 5V supply for external system
• JTAG1: JTAG 10-pin connector
• USB1: micro-USB connector
• EX1, EX2: Female sockets for XBee
• Various test points located throughout the board
113
4.2 Function Blocks
Figure 4.3: Falcon I Autopilot Board Block Diagram.
4.2.1 Microcontroller Unit (MCU)
The Falcon I Autopilot board is built around the STM32F405VGT6, a 32-bit ARM® Cortex® M4 micro-controller running at 168MHz with 192KB RAM, 1024KB Flash, and a floating point unit (FPU). Themicrocontroller includes CAN, I2C, I2S, SPI and UART communication interfaces, 82 I/O and 10 Timers ina LQFP-100 package.
4.2.2 Power Supply
The on-board power supply is split into two 3.3V Low Drop-Out (LDO) voltage regulators. The first voltageregulator (LP2992AIM5-3.3) supplies power to the accelerometer, gyroscope, magnetometer and barometersensors, so as to reduce noise in the measurements, while the second voltage regulator (LM3940IMPX-3.3)supplies power to the rest of the board. The latter is capable of delivering up to 1A.
4.2.3 Inertial Measurement Unit (IMU)
The boards features a 10 Degrees-of-Freedom (DOF) Inertial Measurement Unit (IMU) comprised by a 3-axis accelerometer and 3-axis gyroscope (MPU6000), a 3-axis magnetometer (HMC5883L) and a barometer(MS5611-01BA03).
The accelerometer/gyroscope and the barometer are connected via SPI interface with the MCU, and themagnetometer is connected to the accelerometer/gyroscope via I2C interface.
The accelerometer/gyroscope acts as a master with the magnetometer, pulling the data and saving it locally,allowing the MCU to extract data from these three sensors directly from the MPU6000 via SPI.
114
4.2.4 GPS
The board includes a uBlox MAX-7Q GPS module with a super-capacitor for hot start and antenna matchingcircuitry. The GPS works in the 1575.42Mhz band. The antenna circuitry includes a SAW filter and optionalLow Noise Amplifier (LNA) that can be enabled/disabled by means of on-board 0 ohm jumper resistors.The board also includes uFL and SMA connectors for GPS antenna. The connectors can also be selectedusing 0 ohm jumper resistors.
The default setup includes the SMA connector selected and the LNA disabled, so as to use an active antenna.The use of the on-board LNA is advised only when using passive antennas, since active antennas alreadyhave an LNA.
The GPS communication interface is serial and is connected to the MCU using the UART3 port. The boardalso features jumpers for powering on/off the GPS and connecting/disconnecting the GPS from the UART3port, in case and external GPS is used. Details on these jumper can be found on section 4.2.10.
4.2.5 XBee
The board includes a set of female headers for connecting an XBee RF transceiver module for telemetry.It is important to connect the XBee with the right orientation. Figure 4.4 shows an example of an XBeemodule with SMA connector plugged into the Falcon I Autopilot board. The XBee communication interfaceis serial and is connected to the MCU using the UART2 port.
Figure 4.4: Falcon I Autopilot Board Overview with XBee connected.
4.2.6 Data-Logger
A secondary MCU based on an Atmel AVR 8-bits 16Mhz ATMEGA328P takes care of listening to thetelemetry serial port (UART2) and saving the data in a micro-SD card that can be connected using the micro-SD card holder located in the bottom of the board. The MCU uses a modified version of the “OpenLog”firmware that was officially adapted by paparazziUAV to work with their system. Once the data-loggingfirmware is running there is no need to change any settings, as it starts running and saving data automaticallyafter power up.
The board includes a jumper for powering on/off the data-logger. Details on this jumper can be found onsection 4.2.10.
115
4.2.7 CAN Transceiver
The board features a Microchip MCP2551 CAN transceiver supporting 1M/s operation. The CANH andCANL pins are accessible through connector J5. Details on this connector can be found on section 4.2.9.
4.2.8 Motor/Servo Outputs (PWM)
The Falcon I Autopilot board has eight PWM outputs that can be used to control motors and servos.Outputs run through tristate buffers and each signal can be enabled/disabled using one of two user controlsignals. This functionality is intended for using external kill-switches for motor and servo signals.
PWM outputs and control signals are accessible through connectors J3 and J7 respectively. Details on theseconnectors can be found on section 4.2.9
4.2.9 Connectors
J3 Header
Figure 4.5 shows the location and pinout of the motor/servo headers. There are eight 3-pin headers, eachone with GND, Power and PWM output. The Power pins are all tied together and have 4.8-5.5V range.When an Electronic Speed Controller (ESC) is connected, the power coming from it is used to power theservos.
Each of the 8 PWM outputs can be assigned to one of the two user control switches (as a motor or as aservo), as explained in 4.2.8. This is done via solder jumpers SB6-SB13. Details on these jumpers can befound on section 4.2.10.
Figure 4.5: J3 Header Pinout.
116
J5 HeaderFigure 4.6 shows the location and pinout of the 48-pin expansion header that can used to access communi-cation ports, GPIO, Analog inputs, etc.
Figure 4.6: J5 Header Pinout.J6 HeaderFigure 4.7 shows the location and pinout of the pico-blade connector for UART3. This serial port can beused to connect an external GPS module. If an external GPS module is used, the on-board GPS modulecan be disconnected from power using jumper JP10 and the TX and RX lines can be disconnected from theMCU using jumpers JP2 and JP4. Details on these jumpers can be found on section 4.2.10.
Figure 4.7: J6 Header Pinout.
117
J7 Header
Figure 4.8 shows the location and pinout of the user control switches. This header can also be used withjumper, but is primarily intended to be used with SPST toggle switches, each one connected between pins1-2, 3-4, 5-6, 7-8, and 9-10.
Figure 4.8: J7 Header Pinout.
From top to bottom, the first three pairs are used for enabling/disabling the main power supply, externalpower supply and backup power supply respectively. When the switch (or jumper) in on, those pins areshorted to GND and power is off for the corresponding function. The last two pairs are used for enabling/dis-abling the servo and motor signals respectively. When the switch (or jumper) in on, those pins are shortedto 3.3V, and the servo/motor signals are disabled (PWM signal disabled and output pulled-down to GND).Jumper SB6-SB13 are used to select which PWM outputs are assigned to motors and servos. Details onthese jumpers can be found on section 4.2.10.
118
J9 Header
Figure 4.9 shows the location and pinout of the power header. This header is used for powering the board,and also carries external and backup power supplies as well as power enable signals. Table 4.1 shows adescription of each pin of this header.
Figure 4.9: J9 Header Pinout.
Table 4.1: J9 Header Pinout Description.
Pin Signal Description1 VIN Battery voltage for voltage measurement2 GND Ground3 Backup Power Enable Enable signal for backup power4 Servo Power Backup power for servos5 External Power Enable Enable signal for external power6 External Supply External power (5V) for secondary system7 Main Power Enable Enable signal for main power8 5V Main power (5V)
119
J10 Header
Header J10 includes 3 communication ports: 1 x I2C and 2 x UART.
Figure 4.10: J10 Header Pinout.
J11 Header
Figure 4.11: J11 Header Pinout.
120
J12 Header
Figure 4.12: J12 Header Pinout.
J13 Header
Figure 4.13 shows the location and pinout of the pico-blade connector for UART1. This serial port is usedto connect an RC receiver.
Figure 4.13: J13 Header Pinout.
121
J14 Header
Figure 4.14 shows the location and pinout of the pico-blade connector for UART5. This serial port is usedto connect an RC receiver.
Figure 4.14: J14 Header Pinout.
J15 Header
Figure 4.15 shows the location and pinout of the ISP connector used for programming the data-logger MCUATMEGA328P.
Figure 4.15: J15 Header Pinout.
122
J16 Header
Figure 4.16 shows the location and pinout of the FTDI connector used for communicating with the data-logger MCU ATMEGA328P.
Figure 4.16: J16 Header Pinout.
J17 Header
Figure 4.17 shows the location of the uFL connector for the GPS antenna. This connector can be enabledsetting the jumper resistor R62 instead of R67.
Figure 4.17: J17 Header Pinout.
123
J18 Header
Figure 4.18 shows the location of the SMA connector for the GPS antenna. This connector can be enabledsetting the jumper resistor R67 instead of R62.
Figure 4.18: J18 Header Pinout.
J19 Header
Figure 4.19 shows the location and pinout of the SPI interface connector.
Figure 4.19: J19 Header Pinout.
124
J20 HeaderFigure 4.20 shows the location and pinout of the UART2 connector. This serial port is connected to theXBee module used for telemetry. If an external telemetry module is used, it can be connected to this portwhile leaving the XBee sockets empty.
Figure 4.20: J20 Header Pinout.J21 HeaderFigure 4.21 shows the location and pinout of the UART_LV connector. This UART port has a voltage levelshifter and can be used for communication with an external system that uses a different voltage level (e.g.ODROID Linux mini-PC using 1.8V logic).
Figure 4.21: J21 Header Pinout.
125
This serial port can be connected to UART2 (currently used for the XBee telemetry module) in case thesecondary system only wants to listen to telemetry information; or it can be connected to UART6, whichcan be used as a dedicated serial link. This change can be done using SB1 and SB2 solder jumpers. Detailson these jumpers can be found on section 4.2.10.J22 HeaderFigure 4.22 shows the location and pinout of the External Power connector. This connector provides powerto an external system. Power comes from pin 6 in header J9.
Figure 4.22: J22 Header Pinout.JTAG1 HeaderFigure 4.23 shows the location of the JTAG connector used for programming the STM32F4 MCU.
Figure 4.23: JTAG1 Header Pinout.
126
USB1 Header
Figure 4.24 shows the location of the micro-USB connector.
Figure 4.24: USB1 Header Pinout.
EX1, EX2 Headers
Figure 4.25 shows the location of the XBee socket connectors used for plugging the XBee module. Refer tosection 4.2.5 for correct orientation of the XBee module.
Figure 4.25: EX1, EX2 Headers Pinout.
127
4.2.10 Jumpers
Solder Jumper SB1-SB13
Figure 4.26 shows the location of the solder jumpers SB1 to SB13 on the bottom side of the Falcon I Autopilotboard, and Figure 4.27 shows the three sub-groups of solder jumpers. All solder jumpers have the same 3-pad footprint, where the middle pad is the common pin and the two side pads are the options. To makea connection, place a solder bridge between the middle pad and one of the side pads. Some jumpers comealready with a default connection by means of a small exposed track (no solder mask) joining the middle padwith one of the side pads. If the default option is no desired, the track needs to be cut so as to disconnectthe pads.
Figure 4.26: Location of all solder jumpers.
(a) Solder jumpers SB1-SB2. (b) Solder jumpers SB3-SB5. (c) Solder jumpers SB6-SB13.
Figure 4.27: Solder jumper sub-groups.
Table 4.2 shows the description for each solder jumper and the two options. The location reference for theoptions (Left, Right, Top, Bottom) is based on the orientation of the jumpers shown in Figure 4.27. Notethat SB4 and SB5 have a default option (option 1) already selected by means of a trace connection, so thereis no need to place a solder bridge if this option is desired. If option 2 is desired, the trace must be carefullycut using a sharp blade (e.g. scalpel knife) before placing the solder bridge.
Warning: For SB4 and SB5, do not solder bridge option 2 without removing option 1 default connection.Powering the boards like this might damage it.
128
Table 4.2: Solder jumpers descriptions.
SB Option 1 Option 2 Description1 (Left) UART6_TX (Right) UART2_TX Selects UART (Low Voltage) port (TX pin)2 (Left) UART2_RX (Right) UART6_RX Selects UART (Low Voltage) port (RX pin)3 (Top) 3.3V (Bottom) 5V UART3 Power Rail4 (Top) 3.3V (default) (Bottom) 5V UART1 Power Rail5 (Top) 3.3V (default) (Bottom) 5V UART5 Power Rail6 (Top) Servo Group (Bottom) Motor Group PWM signal enable/disable (Output 1)7 (Top) Servo Group (Bottom) Motor Group PWM signal enable/disable (Output 2)8 (Top) Servo Group (Bottom) Motor Group PWM signal enable/disable (Output 3)9 (Top) Servo Group (Bottom) Motor Group PWM signal enable/disable (Output 4)10 (Top) Servo Group (Bottom) Motor Group PWM signal enable/disable (Output 5)11 (Top) Servo Group (Bottom) Motor Group PWM signal enable/disable (Output 6)12 (Top) Servo Group (Bottom) Motor Group PWM signal enable/disable (Output 7)13 (Top) Servo Group (Bottom) Motor Group PWM signal enable/disable (Output 8)
GPS Signal Jumpers
Figure 4.28 shows the location of the jumpers used for configuring the GPS signal circuitry. There arethree GPS signal setup jumpers used for configuring the signal path. Each jumper consists of an 0402capacitor/resistor that can be soldered vertically or horizontally. The orientation of these componentsdefines the route of the signal. Table 4.3 shows the description for each setup based on these jumpers.
Figure 4.28: Location of GPS signal jumpers.
Table 4.3: GPS signal jumpers descriptions.
Setup DescriptionC84 On and R63 On On-board LNA enabledC86 On and R66 On On-board LNA disabledR62 On uFL antenna connector enabledR67 On SMS antenna connector enabled
129
GPS and Data-logger Power Jumpers
Figure 4.29 shows the location of the two power jumpers. When JP7 is closed the data-logger is poweredand when JP10 in closed the GPS is powered.
Figure 4.29: Location of power jumpers.
GPS Serial Jumpers
Figure 4.30 shows the location of the GPS serial jumpers. These jumpers form a “T” connection for theUART3 port between the on-board GPS module, MCU and J6 header. JP1 and JP3 connect the RX andTX lines respectively of the on-board GPS module to header J6. JP2 and JP4 connect the TX and RX linesrespectively of the MCU to header J6.
Figure 4.30: Location of GPS serial jumpers.
130
During normal operation (using the on-board GPS), all 4 jumpers are closed. There are other two cases:
1. JP1 and JP3 can be left open for using an external GPS module connected directly to J6.
2. JP2 and JP4 can be left open for talking directly to the GPS module from an external system throughJ6. An example of this is connecting to the GPS from a PC in order to configure it.
131
4.3 Board Pictures
Figure 4.31: Falcon I Autopilot Board (Top).
Figure 4.32: Falcon I Autopilot Board (Bottom).
132
4.4 Autopilot Board Schematics
This section contains the following schematics:
• Index
• UAARG Autopilot
• Connectors
• MCU
• IMU
• GPS
• CAN, JTAG, LDO, USB
• Telemetry, I2C
• Switches, UI
• Data Logger
• Power
• Main Power
• Backup Power
• Notes
• Documentation
• Revision History
133
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
AIn
dex.
SchD
oc2.
020
-DEC
-201
51
16En
g. J
orge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
b
U_D
ocum
enta
tion
Doc
umen
tatio
n.Sc
hDoc
U_R
evis
ion
His
tory
Rev
isio
n H
isto
ry.S
chD
oc
UA
AR
G A
utop
ilot.P
rjPc
b2.
0Pr
ojec
t Titl
e:R
evis
ion:
Dat
e:20
-DEC
-201
5
Blo
cks
Page
Nam
e...
......
.....
......
......
......
......
......
......
......
......
......
......
......
......
......
.....
1In
dex
2U
AA
RG
Aut
opilo
t3 4 5 6 7 8 9 18
......
......
......
......
......
......
......
......
......
......
..17
......
......
......
......
......
......
......
......
......
......
..16
......
......
......
......
......
......
......
......
......
......
..1514 23
......
......
......
......
......
......
......
......
......
......
..22
......
......
......
......
......
......
......
......
......
......
..21
......
......
......
......
......
......
......
......
......
......
..20
......
......
......
......
......
......
......
......
......
......
..19
......
......
......
......
......
......
......
......
......
......
..
25...
......
......
......
......
......
......
......
......
......
.....
24...
......
......
......
......
......
......
......
......
......
.....
11
Not
es
10 13
Rev
isio
n H
isto
ry
12
Doc
umen
tatio
n
27...
......
......
......
......
......
......
......
......
......
.....
26...
......
......
......
......
......
......
......
......
......
.....
28...
......
......
......
......
......
......
......
......
......
.....
U_U
AA
RG
Aut
opilo
tU
AA
RG
Aut
opilo
t.Sch
Doc
Dat
a Lo
gger
Tele
met
ry, I
2C
Mai
n Po
wer
U_N
otes
Not
es.S
chD
oc
Uni
vers
ity o
f Alb
erta
Aer
ial R
obot
ics G
roup
MC
U
CA
N, J
TAG
, LD
O, U
SB
IMU
Switc
hes,
UI
Bac
kup
Pow
er
U_P
ower
Pow
er.S
chD
oc
GPS
Pow
er
134
11
22
33
44
55
66
DD
CC
BB
AA
1400
x900
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:Re
v:D
ate:
Shee
t
o
fD
raw
n by
:
UA
RT_I
N
U_D
ata
Logg
erD
ata
Logg
er.S
chD
oc
TP3
TP4
TP1
GN
D
+3V
3+3
V3_
SEN
S
S1 S2 S3 S4 S5 S6 S7 S8
S1 S2 S3 S4 S5 S6 S7 S8
LED
s
Serv
os
BIN
D
BU
FF1_
ENB
UFF
2_EN
U_S
witc
hes,
UI
Switc
hes,
UI.S
chD
oc
PB0
PB2
PB5
PA11
PA12
PC0
PC1
PC3
PC4
PC14
NR
STB
OO
T0Se
rvos
JTA
G
CA
N
I2C
1
I2C
2
LED
s
SPI1
SPI2
UA
RT
USB
U_M
CU
MCU
.Sch
Doc
Serv
os
LED
s
BIN
D
MA
G_I
NT
MPU
_IN
T
GPS
_TX
GPS
_RX
JTA
G
CA
NCA
N_B
US
USB
U_C
AN
, JTA
G, L
DO
, USB
CAN
, JTA
G, L
DO
, USB
.Sch
Doc
JTA
G
CA
N
XB
EE_R
XX
BEE
_TX
LV_R
XLV
_TX
I2C
1
I2C
2
I2C1
_5V
I2C2
_5V
UA
RT_T
XU
ART
_RX
U_T
elem
etry
, I2C
Tele
met
ry, I
2C.S
chD
ocI2
C1
I2C
2
SPI2
UA
RT.
UA
RT2.
RX
UA
RT
UA
RT.
UA
RT2.
TX
UA
RT_T
XU
ART
_RX
LV_T
XLV
_RX
SPI1
I2C1
_5V
I2C2
_5V
CAN
_BU
S
AD
C_1
AD
C_2
AD
C_3
BO
OT0
NR
ST
GPI
O1
GPI
O2
GPI
O3
9.1K
R2
2.2K
R3
GN
D
VB
AT
UA
RT.
UA
RT2.
TX
+5V
UA
RT.
UA
RT1.
TXU
AR
T.U
ART
1.R
XU
AR
T.U
ART
2.TX
UA
RT.
UA
RT2.
RX
UA
RT3_
TX_E
UA
RT3_
RX_E
UA
RT.
UA
RT5.
TXU
AR
T.U
ART
5.R
X
LV_T
XLV
_RX
LV+5
V_A
LT
BO
OT0
NR
ST
GPI
O1
GPI
O2
GPI
O3
AD
C_1
AD
C_2
AD
C_3
UA
RT.
UA
RT5.
TX
SPI1
.MO
SISP
I1.M
ISO
SPI1
.SC
LKSP
I1.C
SSP
I1.D
RDY
CAN
_BU
S.CA
NH
CAN
_BU
S.C
AN
LI2
C1_5
V.S
DA
I2C1
_5V
.SC
L
I2C2
.SD
AI2
C2.S
CL
I2C2
_5V
.SD
AI2
C2_5
V.S
CL
GN
D
GN
D
SET1
_EN
SET2
_EN
12
34
56
78
910
1112
1314
1516
1718
1920
2122
2324
2526
2728
2930
3132
3334
3536
3738
3940
4142
4344
4546
4748
J5 Mai
n Po
rt
GN
D
UA
RT.
UA
RT6.
TXU
AR
T.U
ART
6.R
X
+3V
3+5
V+3
V3
UA
RT.
UA
RT6.
TX
UA
RT.
UA
RT6.
RX
UA
RT.
UA
RT2.
RX
UA
RT.
UA
RT2.
TX3
1
2
SB1
UA
RT_T
X
UA
RT_R
X
UAR
T (L
ow V
olta
ge) S
elec
tion
VB
AT
MH
1
Mou
ntin
g H
oles
3.1
75m
m (1
25m
il) d
rill f
or 4
-40/
M2.
5/M
3 sc
rew
s
FID
1FI
D2
FID
4FI
D5
FID
3FI
D6
Fidu
cial
s
Mec
hani
cal
TP2
+5V
USB
31
2
SB2
MH
2M
H3
MH
4
0R1
MPU
_IN
TM
AG
_IN
T
SPI2
U_I
MU
IMU
.Sch
Doc
GPS
_RX
GPS
_TX
U_G
PSG
PS.S
chD
oc
GN
D
SET1
_EN
SET2
_EN
+3V
3
12
34
56
78
910
J7 Switc
hes
BC
KP_
EN
VIN
_REF
MA
IN_E
NA
LT_E
NB
CK
P_EN
M_E
XT_
EN
12
34
56
78
J9 Pow
er
S1 S2 S3 S4 S5 S6 S7 S8 M_E
XT_
EN
I2C
2
I2C2
_5V
UA
RT
GPS
_TX
GPS
_RX
SPI1
LV_T
XLV
_RX
UA
RT3_
RX_E
UA
RT3_
TX_E
U_C
onne
ctor
sCo
nnec
tors
.Sch
Doc
I2C
2
M_E
XT_
EN
UA
RT1_
VC
C
UA
RT5_
VC
CU
ART
3_V
CC
UA
AR
G A
utop
ilot.S
chD
oc2.
020
-DEC
-201
52
16En
g. Jo
rge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
SERV
O_P
WR
GN
D
+5V
+5V
_ALT
VIN_REFM
AIN
_EN
ALT
_EN
UA
RT3_
RX_E
UA
RT3_
TX_E
COFID1
COFI
D2COFID3
COFID4
COFI
D5CO
FID6
PIJ501
PIJ502
PIJ503
PIJ504
PIJ505
PIJ506
PIJ507
PIJ508
PIJ509
PIJ5010
PIJ5011
PIJ5012
PIJ5
013
PIJ5014
PIJ5015
PIJ5016
PIJ5017
PIJ5018
PIJ5019
PIJ5020
PIJ5021
PIJ5022
PIJ5023
PIJ5024
PIJ5025
PIJ5026
PIJ5
027
PIJ5028
PIJ5029
PIJ5030
PIJ5
031
PIJ5032
PIJ5
033
PIJ5034
PIJ5035
PIJ5036
PIJ5
037
PIJ5038
PIJ5039
PIJ5040
PIJ5041
PIJ5042
PIJ5043
PIJ5044
PIJ5045
PIJ5046
PIJ5047
PIJ5048
COJ5
PIJ701
PIJ702
PIJ703
PIJ704
PIJ705
PIJ706
PIJ707
PIJ708
PIJ709
PIJ7010
COJ7
PIJ901
PIJ902
PIJ903
PIJ904
PIJ905
PIJ906
PIJ907
PIJ908
COJ9
PIMH101COMH1
PIMH201COMH2
PIMH301COMH
3
PIMH401COMH4
PIR101PIR102 COR1
PIR201PIR202 COR2
PIR301PIR302 COR3
PISB101
PISB102PISB103
COSB1
PISB201
PISB202PISB203
COSB2
PITP101 COTP
1
PITP201 COTP
2
PITP301 COTP3
PITP401 COTP4
PIJ503
PIJ504
PIJ707
PIJ709 PITP301
PITP401
PIJ501
PIJ502
PIJ908
PITP201
PIJ5047
PIJ906
PIJ5019
NLAD
C01
PIJ5021
NLAD
C02
PIJ5023
NLAD
C03
PIJ704
PIJ905
NLALT0EN
PIJ706
PIJ903
NLBC
KP0E
N
NLBI
ND
PIJ509
NLBOOT0
NLCA
NNL
CAN
PIJ5035
NLCA
N0BU
S
NLCAN0BUS0CANH
PIJ5
037
NLCA
N0BU
S
NLCAN0BUS0CANL
PIJ505
PIJ506
PIJ5048
PIJ701
PIJ703
PIJ705
PIJ902
PIR301
PITP101
PIJ5011
NLGP
IO1
PIJ5
013
NLGP
IO2
PIJ5015
NLGP
IO3
NLGPS0RX
NLGPS0TX
NLI2
C1NL
I2C1
PIJ5041
NLI2
C105
V
NLI2
C105
V0SC
LPIJ5039
NLI2
C105
VNL
I2C1
05V0
SDA
PIJ5036
NLI2
C2
NLI2
C20S
CLPIJ5034
NLI2
C2
NLI2C20SDA
PIJ5040
NLI2
C205
VNL
I2C2
05V0
SCL
PIJ5038
NLI2
C205
V
NLI2
C205
V0SD
A
NLJT
AGNL
JTAG
NLJT
AGNL
JTAG
NLJT
AG
NLLE
DsNL
LEDs
NLLE
DsNL
LEDs
NLLE
Ds
PIJ5046
PIJ5042
NLLV0RX
PIJ5044
NLLV0TX
PIJ5043
NLM0EXT0EN
NLMA
G0IN
T
PIJ702
PIJ907
NLMA
IN0E
N
NLMP
U0IN
T
PIMH101PIMH201
PIMH301PIMH401
PIR101PIR201 PIR302
PIJ507
NLNR
STNLS
1NLS
2NLS
3NLS
4NLS
5NLS
6NLS
7NLS
8
PIJ904
NLSe
rvos
NLSe
rvos
NLSe
rvos
NLSe
rvos
NLSe
rvos
NLSe
rvos
NLSe
rvos
NLSe
rvos
PIJ708
NLSE
T10E
N
PIJ7010
NLSE
T20E
N
PIJ5
031
NLSP
I1
NLSP
I10C
S
PIJ5
033
NLSP
I1
NLSPI10DRDY
PIJ5
027
NLSP
I1
NLSPI10MISO
PIJ5025
NLSP
I1
NLSPI10MOSI
PIJ5029
NLSP
I1
NLSPI10SCLK
NLSP
I2NL
SPI2
NLSP
I2NL
SPI2
NLSP
I2
PIJ508
PIJ5024
NLUART30RX0E
PIJ5022
NLUART30TX0E
PIJ5010
PIJ5012
PIJ5016
NLUA
RT
NLUA
RT0U
ART1
0RX
PIJ5014
NLUA
RT
NLUART0UART10TX
PIJ5020
PISB203
NLUA
RT
NLUA
RT0U
ART2
0RX
PIJ5018
PISB103
NLUA
RT
NLUART0UART20TX
NLUA
RTNL
UART
PIJ5028
NLUA
RT
NLUA
RT0U
ART5
0RX
PIJ5017
PIJ5026
NLUA
RT
NLUART0UART50TX
PIJ5032
PISB201
NLUA
RT
NLUART0UART60RX
PIJ5030
PISB101
NLUA
RT
NLUART0UART60TX
PISB202
NLUA
RT0R
X
PISB102
NLUA
RT0T
X
NLUS
BNL
USB
NLUS
B
PIJ5045
PIR102
NLVBAT
PIJ901
PIR202NLVIN0REF
135
11
22
33
44
55
66
DD
CC
BB
AA
1400
x900
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:Re
v:D
ate:
Shee
t
o
fD
raw
n by
:
S_G
ND
SERV
O_P
WR
S1
S_G
ND
SERV
O_P
WR
S2
S_G
ND
SERV
O_P
WR
S3
S_G
ND
SERV
O_P
WR
S4
S_G
ND
SERV
O_P
WR
S5
S_G
ND
SERV
O_P
WR
S6
S_G
ND
SERV
O_P
WR
S7
S_G
ND
SERV
O_P
WR
S8
Serv
os/M
otor
(s)
+3V
3+5
V
JP1
UA
RT3_
TX_E
UA
RT3_
RX_E
GPS
_TX
GPS
_RX
10K
DN
PR
4
GN
D
+3V
3+5
V
UA
RT1_
TX10
KR
7
UA
RT1_
RX
UA
RT5_
TXU
ART
5_R
X
31
2
SB4
UAR
T1 &
UAR
T5 (R
C)
UAR
T3 (G
PS)
GN
D
+3V
3
UA
RT1_
VC
C
GN
D
+3V
3+5
V
10K
R8
UA
RT5_
VC
C
I2C2
31
2
SB3
31
2
SB5
1 2 3
J3A 4 5 6
J3B 7 8 9
J3C 10 11 12
J3D 13 14 15
J3E 16 17 18
J3F 19 20 21
J3G 22 23 24
J3H
1 2 3 4
J10A
UA
RT1
UA
RT5
I2C
2
Serv
o 1
Serv
o 2
Serv
o 3
Serv
o 4
Serv
o 5
Serv
o 6
Serv
o 7
Serv
o 8
5 6 7 8
J10B
9 10 11 12
J10C
81
1KRN1A
72
1KRN1B
63
1KRN1C
54
1KRN1D
81
1KRN2A
72
1KRN2B
63
1KRN2C
54
1KRN2D
1 2 3 4
J13
PB-U
AR
T1
1 2 3 4
J14
PB-U
AR
T5
1 2 3 4
J6 PB-U
AR
T3
iPC
B R
ule
iPC
B R
ule
Mot
or(s
) Ext
erna
l Shu
tdow
n
C1
B1
E1MMDT5551-7-F
Q1A
GN
D C2
B2
E2MMDT5551-7-F
Q1B
GN
D
0R5
0R9
S8_O
UT
S7_O
UT
S8_O
UT
S7_O
UT
M_E
XT_
EN1KR
6
1KR10
S_G
ND
GN
D
JP3
JP2
JP4
Con
nect
ors.S
chD
oc2.
020
-DEC
-201
53
16En
g. Jo
rge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
S1 S2 S3 S4 S5 S6 S7 S8
M_E
XT_
ENSD
ASC
L
I2C
2
I2C
2
SDA
SCL
I2C2
_5V
I2C2
_5V
I2C2
_SC
LI2
C2_S
DA
I2C2
_5V
_SC
LI2
C2_5
V_S
DA
UA
RT1_
TXU
ART
1_R
XU
ART
5_TX
UA
RT5_
RX
TX RX
UART2
UA
RT
TX RX
UART3TX RX
UART1 TX RX
UART5 TX RX
UART6
UA
RT1
UA
RT2
UA
RT3
UA
RT5
UA
RT6
UA
RT
UA
RT1_
TXU
ART
1_R
X
UA
RT5_
TXU
ART
5_R
X
UA
RT3_
TX
UA
RT3_
RX
UA
RT3_
TXU
ART
3_R
X
GPS
_TX
GPS
_RX
I2C2
_SC
LI2
C2_S
DA
GN
D
+3V
3 I2C2
_SC
LI2
C2_S
DA
1 2 3 4
J12
PB-I2
C2_5
VG
ND
+5V
I2C2
_5V
_SC
LI2
C2_5
V_S
DA
1 2 3 4 5 6 7
J19
PB-S
PI1
SCLK
MO
SIM
ISO CS
DRD
Y
SPI1
SPI1
SPI1
_MO
SISP
I1_C
SSP
I1_D
RDY
SPI1
_MIS
OSP
I1_S
CLK
+3V
3
GN
D
1 2 3 4
J20
PB-U
AR
T2
UA
RT2_
TXU
ART
2_R
X
GN
DG
ND
GN
D
+3V
3
UA
RT2_
TXU
ART
2_R
X
1 2 3 4 5
J21
PB-U
AR
T_LV
LV_T
XLV
_RX
M_E
XT_
EN
LV
GN
D GN
D
LV_T
XLV
_RX
UAR
T (E
xter
nal V
olta
ge L
evel
)
SERV
O_P
WR
SERV
O_P
WR
SPI1
_MO
SI
SPI1
_CS
SPI1
_DRD
Y
SPI1
_MIS
OSP
I1_S
CLK
UA
RT1_
VC
CU
ART
5_V
CC
UA
RT3_
VC
C
1 2 3 4
J11
PB-I2
C2
UAR
T2 (T
elem
etry
)
SPI1
UA
RT3_
TX_E
UA
RT3_
TX_E
UA
RT3_
RX_E
UA
RT3_
RX_E
1 2
J22
Ext.
Pow
er
Exte
rnal
Pow
erGN
D
+5V
_ALT
PIJ301
PIJ302
PIJ303
COJ3
A
PIJ304
PIJ305
PIJ306
COJ3
B
PIJ307
PIJ308
PIJ309
COJ3
C
PIJ3010
PIJ3011
PIJ3012CO
J3D
PIJ3013
PIJ3014
PIJ3015CO
J3E
PIJ3016
PIJ3017
PIJ3018CO
J3F
PIJ3019
PIJ3020
PIJ3021CO
J3G
PIJ3022
PIJ3023
PIJ3024CO
J3H
PIJ601
PIJ602
PIJ603
PIJ604COJ6
PIJ1001
PIJ1002
PIJ1003
PIJ1004
COJ1
0APIJ1005
PIJ1006
PIJ1007
PIJ1008
COJ1
0B
PIJ1009
PIJ10010
PIJ10011
PIJ10012
COJ1
0C
PIJ1101
PIJ1102
PIJ1103
PIJ1104
COJ1
1
PIJ1201
PIJ1202
PIJ1203
PIJ1204
COJ1
2
PIJ1301
PIJ1302
PIJ1303
PIJ1304
COJ1
3
PIJ1401
PIJ1402
PIJ1403
PIJ1404
COJ1
4
PIJ1901
PIJ1902
PIJ1903
PIJ1904
PIJ1905
PIJ1906
PIJ1907
COJ19
PIJ2001
PIJ2002
PIJ2003
PIJ2004
COJ20
PIJ2101
PIJ2102
PIJ2103
PIJ2104
PIJ2105
COJ21
PIJ2201
PIJ2202
COJ22
PIJP
101
PIJP
102
COJP1
PIJP
201
PIJP
202COJP2
PIJP
301
PIJP
302
COJP3
PIJP
401
PIJP
402 COJP4
PIQ10B1
PIQ10C1 PIQ10E1CO
Q1A
PIQ10B2
PIQ10C2 PIQ10E2
COQ1
B
PIR4
01PI
R402
COR4
PIR501
PIR5
02COR5
PIR601
PIR6
02COR6
PIR7
01PIR702
COR7
PIR8
01PI
R802
COR8
PIR901
PIR9
02COR9
PIR1
001
PIR1002COR
10
PIRN101
PIRN
108
CORN
1A
PIRN102
PIRN10
7
CORN
1B
PIRN103
PIRN
106
CORN
1C
PIRN104
PIRN10
5
CORN
1D
PIRN201
PIRN20
8
CORN
2A
PIRN202
PIRN20
7
CORN
2B
PIRN203
PIRN20
6
CORN
2C
PIRN204
PIRN20
5
CORN
2D
PISB301
PISB302PISB303
COSB3
PISB401
PISB402PISB403
COSB4
PISB501
PISB502PISB503
COSB5
PIJ1009
PIJ1101
PIJ1902
PIJ2001
PISB301
PISB401
PISB501
PIJ1201
PISB303
PISB403
PISB503
PIJ2201
PIJ303
PIJ306
PIJ309
PIJ3012
PIJ3015
PIJ3018
PIJ3021
PIJ3024
PIJ604
PIJ1004
PIJ1008
PIJ10012
PIJ1104
PIJ1204
PIJ1304
PIJ1404
PIJ1901
PIJ2004
PIJ2105
PIJ2202
PIQ10E1 PIQ10E2
PIJP
401
NLGPS0RX
POGPS0RX
PIJP
201
NLGPS0TX
POGPS0TX
PIJ1203
NLI2C205V0SCL
POI2
C205
V
PIJ1202
NLI2C205V0SDA
POI2
C205
V
PIJ10011
PIJ1103
NLI2C20SCL
POI2C2
PIJ10010
PIJ1102
NLI2
C20S
DAPOI2C2
PIJ2101
PIJ2103NL
LV0RX
POLV0RX
PIJ2102NLLV0TX
POLV0TX
PIJ2104
PIR6
02
PIR1002
NLM0EXT0EN
POM0EXT0EN
POUART
POUART
PIJ301
PIRN10
5
PIJ304
PIRN
106
PIJ307
PIRN10
7
PIJ3010
PIRN
108
PIJ3013
PIRN20
5
PIJ3016
PIRN20
6
PIQ10B1
PIR601
PIQ10B2
PIR1
001
PIQ10C1PIR501
PIQ10C2PIR901
PIRN104
NLS1
POS1
PIRN103
NLS2
POS2
PIRN102
NLS3
POS3
PIRN101
NLS4
POS4
PIRN204
NLS5
POS5
PIRN203
NLS6
POS6
PIRN202
NLS7
POS7
PIJ3019
PIR5
02
PIRN20
7NLS70OUT
PIRN201
NLS8
POS8
PIJ3022
PIR9
02
PIRN20
8NLS80OUT
PIJ302
PIJ305
PIJ308
PIJ3011
PIJ3014
PIJ3017
PIJ3020
PIJ3023
NLSERVO0PWR
PIJ1906
NLSP
I10C
SPOSPI1
PIJ1907
NLSP
I10D
RDY
POSPI1
PIJ1904
NLSPI10MISO
POSPI1
PIJ1903
NLSPI10MOSI
POSPI1
PIJ1905
NLSPI10SCLK
POSPI1
PIJ1003
PIJ1303
PIR7
01
NLUART10RX
POUART
PIJ1002
PIJ1302
NLUART10TX
POUART
PIJ1001
PIJ1301
PIR702
PISB402NLUART10VCC
PIJ2003
NLUART20RX
POUART
PIJ2002
NLUART20TX
POUART
PIJP
101
NLUART30RX
POUART
PIJ603
PIJP
102
PIJP
202
PIR4
01
NLUART30RX0E
POUART30RX0E
PIJP
301
NLUART30TX
POUART
PIJ602
PIJP
302
PIJP
402
NLUART30TX0E
POUART30TX0E
PIJ601
PIR4
02
PISB302
PIJ1007
PIJ1403
PIR8
01
NLUART50RX
POUART
PIJ1006
PIJ1402
NLUART50TX
POUART
PIJ1005
PIJ1401
PIR8
02
PISB502NLUART50VCC
POGPS0RX
POGPS0TX
POI2C2
POI2C20SCL
POI2C20SDA
POI2
C205
VPO
I2C2
05V0
SCL
POI2
C205
V0SD
A
POLV0RX
POLV0TX
POM0EXT0EN
POS1
POS2
POS3
POS4
POS5
POS6
POS7
POS8
POSPI1
POSP
I10C
SPO
SPI1
0DRD
YPO
SPI1
0MIS
OPO
SPI1
0MOS
IPO
SPI1
0SCL
K
POUART
POUART30RX0E
POUART30TX0E
POUART0UART1
POUART
0UART1
0RXPOU
ART0UA
RT10TX
POUART0UART2
POUART
0UART2
0RXPOU
ART0UA
RT20TX
POUART0UART3
POUART
0UART3
0RXPOU
ART0UA
RT30TX
POUART0UART5
POUART
0UART5
0RXPOU
ART0UA
RT50TX
POUART0UART6
POUART
0UART6
0RXPOU
ART0UA
RT60TX
136
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
AM
CU
.Sch
Doc
2.0
20-D
EC-2
015
416
Eng.
Jor
ge M
arin
(mar
inm
ar@
ualb
erta
.ca)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
PA0-
WK
UP(
PA0)
14
PA1
15
PA2
16
PA3
17
PA4
20
PA5
21
PA6
22
PA7
23
PB0
26
PB1
27
PB2-
BO
OT1
(PB
2)28
PB10
29
PB11
30
PB12
33
PB13
34
PB14
35
PB15
36
PA8
41
PA9
42
PA10
43
PA11
44
PA12
45
PA13
(JTM
S-SW
DIO
)46
PA14
(JTC
K-S
WC
LK)
49
PA15
(JTD
I)50
PB3(
JTD
O/T
RA
CES
WO
)55
PB4(
NJT
RST
)56
PB5
57
PB6
58
PB7
59
PB8
61
PB9
62
PC13
2
PC14
-OSC
32_I
N(P
C14
)3
PC15
-OSC
32_O
UT(
PC15
)4
PC0
8
PC1
9
PC2
10
PC3
11
PC4
24
PC5
25
PC6
37
PC7
38
PC8
39
PC9
40
PC10
51
PC11
52
PC12
53
PD2
54
PH0-
OSC
_IN
(PH
0)5
PH1-
OSC
_OU
T(PH
1)6
NR
ST7
VC
AP_
131
VC
AP_
247
BO
OT0
60
U1A
STM
32F4
05R
GT6
VB
AT
1
VSS
A12
VD
DA
13
VSS
18
VD
D19
VD
D32
VD
D48
VSS
63V
DD
64
U1B
STM
32F4
05R
GT6
2.2u
FC
3
GN
D
GN
D
+3V
3 4.7u
F
C6
GN
D
1uF
C11
GN
D
+3V
3
BLM
15A
X10
2SN
1D
L1
+3V
3
PB0
PB0
PB2
PB5
PB1
PB2
PB3
PB4
PB5
PB6
PB7
PB8
PB9
PB10
PB11
PB12
PB13
PB14
PB15
PA0
PA1
PA2
PA3
PA4
PA5
PA6
PA7
PA8
PA9
PA10
PA11
PA12
PA13
PA14
PA15
PC0
PC1
PC2
PC3
PC4
PC5
PC6
PC7
PC8
PC9
PC10
PC11
PC12
PC13
PC14
PC15
PA11
PA12
PC0
PC1
PC3
PC4
PC14
PD2
1 2
Y1
XTA
L_1
10pF
C2
GN
D
10K
R12
GN
D
+3V
3
NR
ST
NR
ST
BO
OT0
BO
OT0
PH0-
OSC
_IN
PH0-
OSC
_OU
T
GN
D
SPI2
_MO
SISP
I2_M
ISO
SPI2
_SC
LKSP
I2_C
S_M
PU
JTA
G_T
DI
JTA
G_T
CK
JTA
G_T
MS
JTA
G_R
ST
JTA
G_T
DO
MPU
_IN
T
SPI2
_CS_
BA
RO
MA
G_I
NT
UA
RT3
_TX
UA
RT3
_RX
CA
N_R
XC
AN
_TX
VD
DA
UA
RT2
_TX
UA
RT2
_RX
LED
4
LED
3
LED
1
LED
5
LED
2
BIN
D_B
UTT
ON
SER
VO
_7/I2
C1_
SCL
SER
VO
_8/I2
C1_
SDA
I2C
2_SC
LI2
C2_
SDA
UA
RT5
_RX
SPI1
_DR
DY
PC12
/UA
RT5
_TX
SER
VO
_4SE
RV
O_3
SER
VO
_2SE
RV
O_1
VB
AT_
MEA
SA
DC
_1
AD
C_3
AD
C_2
UA
RT1
_RX
UA
RT1
_TX
/USB
_VB
US
SPI1
_CS
SPI1
_MO
SISP
I1_M
ISO
SPI1
_SC
LK
SER
VO
_6SE
RV
O_5
100n
F
C7
100n
F
C8
100n
F
C9
100n
F
C10
100n
F
C12
10K
R11
2.2u
FC
4
2.2u
FC
5
10pF
C1
Serv
o1Se
rvo2
Serv
o3Se
rvo4
Serv
o5Se
rvo6
Serv
o7Se
rvo8
Serv
osPC
6PC
7PC
8PC
9PA
0PA
1PB
6PB
7
Serv
os
JTA
G
PA13
PA14
PA15
PB3
NR
ST
TMS
TCK
TDO
TDI
RST
JTA
G
TX RXC
AN
CA
NPB
8PB
9
SDA
SCL
I2C
1
I2C
1
SDA
SCL
I2C
2
I2C
2PB
10PB
11
PB7
PB6
LED
s
LED
1LE
D2
LED
3LE
D4
LED
5
LED
s
PB4
PC15
PC5
PC2
PA8
SCLK
MO
SIM
ISO
CS
DR
DY
SPI1
SPI1
PA7
PA6
PA5
PA4
SCLK
MO
SIM
ISO
CS_
MPU
CS_
BA
RO
SPI2
SPI2
PB13
PB15
PB14
PB12
PB1
PC13
PB0
PA11
/USB
_DN
PA12
/USB
_DP
VB
US
D_N
D_P
USB
USB
TX RX
UART2
UA
RT
PA2
PA3
TX RX
UART3
PC10
PC11
TX RX
UART1
PA9
PA10
TX RX
UART5
TX RX
UART6
PD2
PC12
PC6
PC7
UA
RT1
UA
RT2
UA
RT3
UA
RT5
UA
RT6
UA
RT
PA9
PA11
PA12
PIC1
01PIC102
COC1
PIC2
01PIC202
COC2
PIC301PIC302COC3
PIC401PIC402COC
4
PIC501PIC502COC
5
PIC601PIC602COC6
PIC701 PIC702COC7
PIC801 PIC802COC8
PIC901 PIC902COC9
PIC1001 PIC1002COC1
0
PIC1101PIC1102COC1
1PIC1201 PIC1202CO
C12
PIL1
01PI
L102
COL1
PIR1101PIR1102 COR1
1
PIR1201PIR1202 COR1
2PIU102
PIU103
PIU104
PIU105
PIU106
PIU107
PIU108
PIU109
PIU1010
PIU1011
PIU1014
PIU1015
PIU1016
PIU1017
PIU1020
PIU1021
PIU1022
PIU1023
PIU1024
PIU1025
PIU1026
PIU1027
PIU1028
PIU1029
PIU1030
PIU1031
PIU1033
PIU1034
PIU1035
PIU1036
PIU1037
PIU1038
PIU1039
PIU1040
PIU1041
PIU1042
PIU1043
PIU1044
PIU1045
PIU1046
PIU1047
PIU1049
PIU1050
PIU1051
PIU1052
PIU1053
PIU1054
PIU1055
PIU1056
PIU1057
PIU1058
PIU1059
PIU1060
PIU1061
PIU1062
COU1
A
PIU101
PIU1012
PIU1013
PIU1018
PIU1019
PIU1032
PIU1048
PIU1063
PIU1064CO
U1B
PIY101
PIY102
PIY103
PIY104
COY1
PIC602PIC701
PIC801PIC901
PIC1001
PIL1
01
PIR1202
PIU101
PIU1019
PIU1032
PIU1048
PIU1064
PIR1102PIU1060
NLBOOT0
POBOOT0
PIC1
01
PIC2
01
PIC301PIC401
PIC501
PIC601PIC702
PIC802PIC902
PIC1002
PIC1101PIC1202
PIR1101
PIU1012
PIU1018
PIU1063
PIY103
PIY104
PIC302PIU1047
PIC402
PIU1031
PIC502PIR1201
PIU107
NLNR
ST
POJTAG
PONRST
PIU1014
NLPA0
POServos
PIU1015
NLPA1
POServos
PIU1016
NLPA2
POUART
PIU1017
NLPA3
POUART
PIU1020
NLPA4
POSPI1
PIU1021
NLPA5
POSPI1
PIU1022
NLPA6
POSPI1
PIU1023
NLPA7
POSPI1
PIU1041
NLPA8
POLEDs
PIU1042
NLPA9
POUART
POUSB
PIU1043
NLPA10
POUART
PIU1044
NLPA11
POPA11
POUSB
PIU1045
NLPA12
POPA12
POUSB
PIU1046
NLPA13
POJTAG
PIU1049
NLPA14
POJTAG
PIU1050
NLPA15
POJTAG
PIU1026
NLPB0
POPB0
PIU1027
NLPB1
POSPI1
PIU1028
NLPB2
POPB2
PIU1055
NLPB3
POJTAG
PIU1056
NLPB4
POLEDs
PIU1057
NLPB
5POPB5
PIU1058
NLPB6
POI2C1
POServos
PIU1059
NLPB7
POI2C1
POServos
PIU1061
NLPB8
POCAN
PIU1062
NLPB9
POCAN
PIU1029
NLPB
10POI2C2
PIU1030
NLPB
11POI2C2
PIU1033
NLPB
12POSPI2
PIU1034
NLPB
13
POSPI2
PIU1035
NLPB
14POSPI2
PIU1036
NLPB
15POSPI2
PIU108
NLPC0
POPC0
PIU109
NLPC1
POPC1
PIU1010
NLPC2
POLEDs
PIU1011
NLPC3
POPC3
PIU1024
NLPC4
POPC4
PIU1025
NLPC5
POLEDs
PIU1037
NLPC6
POServos
POUART
PIU1038
NLPC7
POServos
POUART
PIU1039
NLPC8
POServos
PIU1040
NLPC9
POServos
PIU1051
NLPC
10POUART
PIU1052
NLPC
11POUART
PIU1053
NLPC
12
POUART
PIU102
NLPC
13
POSPI2
PIU103
NLPC
14POPC14
PIU104
NLPC15
POLEDs
PIU1054
NLPD2
POUART
PIC102
PIU105
PIY101
NLPH
00OS
C0IN
PIC202
PIU106
PIY102
NLPH00OSC0OUT
PIC1102PIC1201
PIL1
02
PIU1013
NLVD
DA
POBOOT0
POCAN
POCAN0RX
POCAN0TX
POI2C1
POI2C10SCL
POI2C10SDA
POI2C2
POI2C20SCL
POI2C20SDA
POJTAG
POJTAG0RST
POJTAG0TCK
POJTAG0TDI
POJTAG0TDO
POJTAG0TMS
POLEDS
POLE
DS0L
ED1
POLE
DS0L
ED2
POLE
DS0L
ED3
POLE
DS0L
ED4
POLE
DS0L
ED5
PONRST
POPA11
POPA12
POPB0
POPB2
POPB5
POPC0
POPC1
POPC3
POPC4
POPC14
POSERVOS
POSERV
OS0SER
VO1POS
ERVOS0
SERVO2
POSERV
OS0SER
VO3POS
ERVOS0
SERVO4
POSERV
OS0SER
VO5POS
ERVOS0
SERVO6
POSERV
OS0SER
VO7POS
ERVOS0
SERVO8
POSPI1
POSPI10CS
POSP
I10D
RDY
POSP
I10M
ISO
POSP
I10M
OSI
POSP
I10S
CLK
POSPI2
POSPI20CS0BARO
POSPI20CS0MPU
POSP
I20M
ISO
POSP
I20M
OSI
POSP
I20S
CLK
POUART
POUA
RT0U
ART1
POUART
0UART1
0RXPOU
ART0UA
RT10TX
POUA
RT0U
ART2
POUART
0UART2
0RXPOU
ART0UA
RT20TX
POUA
RT0U
ART3
POUART
0UART3
0RXPOU
ART0UA
RT30TX
POUA
RT0U
ART5
POUART
0UART5
0RXPOU
ART0UA
RT50TX
POUA
RT0U
ART6
POUART
0UART6
0RXPOU
ART0UA
RT60TX
POUSB
POUSB0D0N
POUSB0D0P
POUSB0VBUS
137
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
AIM
U.S
chD
oc2.
020
-DEC
-201
55
16En
g. J
orge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
MPU
-600
0
CLK
IN1
NC
2
RSVD21
SDA/SDI24
SCL/SCLK23
RSVD22
VD
D13
GN
D18
AUX_CL 7
AD0/SDO 9
INT 12
CS 8
FSYNC 11REGOUT 10
NC
3
NC
4
NC
5
AU
X_D
A6
NC
14N
C15
NC
16N
C17
CPOUT20
RSVD19
GND EP
U3
2.2n
FC
14
GN
D
GN
D
GN
D
GN
DG
ND
GN
DM
ISO
CS_
MPU
MO
SISC
LK
MPU
_SC
L
MPU
_SD
A
MPU
_IN
T
Acc
eler
omet
er &
Gyr
osco
pe
MS5
611-
01B
A03
VD
D1
CSB
5
PS2
SCLK
8SD
A/S
DI
7
GN
D3
SDO
6
CSB
4
U2
GN
DG
ND
CS_
BA
RO
Bar
omet
er
SCL
1
NC
5V
DD
2
SETP
8
NC
7
NC
3
NC
6S1
4
GN
D9
C1
10
GN
D11
SETC
12
VD
DIO
13
NC
14
DR
DY
15
SDA
16
U4
HM
C58
83L
220n
FC
18
MA
G_I
NT
GN
DG
ND
MPU
_SC
LM
PU_S
DA
GN
D
Mag
neto
met
er
2.2K
R13
MPU
_SC
LM
PU_S
DA
100n
FC
16
100n
FC
15
100n
FC
19
100n
FC
13
10K
R15
2.2K
R14
+3V
3_SE
NS
+3V
3_SE
NS
+3V
3_SE
NS
+3V
3_SE
NS
+3V
3_SE
NS
SCLK
MO
SIM
ISO
CS_
MPU
CS_
BA
RO
SPI2
SPI2
CS_
MPU
MIS
OM
OSI
SCLK
MIS
OM
OSI
SCLK
CS_
BA
RO
TP5
TP6
TP7
TP8
TP9
4.7u
FC
17
GN
D
PIC1301 PIC1302CO
C13
PIC1401PIC1402CO
C14
PIC1501 PIC1502CO
C15
PIC1601 PIC1602CO
C16
PIC1701 PIC1702CO
C17
PIC1801 PIC1802CO
C18
PIC1901PIC1902CO
C19
PIR1301PIR1302 COR1
3
PIR1401PIR1402 COR1
4
PIR1501PIR1502 COR1
5
PITP
501
COTP5
PITP
601
COTP
6
PITP
701
COTP7
PITP
801
COTP
8PI
TP90
1CO
TP9
PIU201
PIU202
PIU203
PIU204
PIU205
PIU206
PIU207
PIU208
COU2
PIU301
PIU302
PIU303
PIU304
PIU305
PIU306
PIU307PIU308PIU309PIU3010
PIU3011PIU3012
PIU3013
PIU3014
PIU3015
PIU3016
PIU3017
PIU3018
PIU3019PIU3020PIU3021PIU3022PIU3023PIU3024
PIU30EP
COU3
PIU401
PIU402
PIU403
PIU404
PIU405
PIU406
PIU407
PIU408
PIU409
PIU4010
PIU4011
PIU4012
PIU4013
PIU4014
PIU4015
PIU4016
COU4
PIC1301
PIC1501
PIC1601
PIR1302PIR1402
PIR1502
PIU201
PIU3013
PIU402
PIU404
PIU4013
PITP
901
PIU204
PIU205
NLCS0BARO
POSPI2
PITP
801
PIU308
NLCS
0MPU
POSPI2
PIC1302
PIC1402
PIC1502
PIC1602
PIC1702
PIC1901
PIU202
PIU203
PIU301
PIU3011
PIU3018
PIU30EP
PIU409
PIU4011
PIU4015
POMAG0INT
PITP
601
PIU206
PIU309
NLMI
SO
POSPI2
PITP
701
PIU207
PIU3024
NLMO
SI
POSPI2
PIR1501
PIU3012
POMPU0INT
PIR1401
PIU307PIU401
NLMPU0SCL
PIR1301PIU306
PIU4016
NLMPU0SDA
PIC1401 PIU3020
PIC1701PIU4010
PIC1801PIU408
PIC1802PIU4012
PIC1902
PIU3010
PIU302
PIU303
PIU304
PIU305
PIU3014
PIU3015
PIU3016
PIU3017
PIU3019PIU3021PIU3022
PIU403
PIU405
PIU406
PIU407
PIU4014
PITP
501
PIU208
PIU3023
NLSC
LK
POSPI2
POMAG0INT
POMPU0INT
POSPI2
POSPI20CS0BARO
POSPI20CS0MPU
POSP
I20M
ISO
POSP
I20M
OSI
POSP
I20S
CLK
138
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
AG
PS.S
chD
oc2.
020
-DEC
-201
56
16En
g. J
orge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
MA
X-7
Q
GN
D1
TXD
2
RX
D3
TIM
EPU
LSE
4EX
TIN
T5
V_B
CK
P6
VC
C_I
O7
VC
C8
RES
ET9
GN
D10
RF_
IN11
GN
D12
AN
T_O
N13
VC
C_R
F14
RES
ERV
ED15
SDA
16
SCL
17
RES
ERV
ED18
U23
GN
D
GPS
_RX
GPS
_TX
GPS
_RX
GPS
_TX
+3V
3_G
PS
11m
FC
87
LQW
15A
N6N
8H00
D
L6 6.8n
H
1
GND GND
J17
U.F
L
GN
D
GN
D
1N41
48W
X-T
PD
4
100
R61
+3V
3_G
PS
GN
D100n
FC
83
+3V
3_G
PS
+3V
3
1
GND GND
J18
SMA
GN
D
MA
X26
59EL
T+T
RFI
N3
GN
D2
VC
C4
RFO
UT
6SH
DN
5G
ND
1
U25
470p
F
C89
33nF
C88
GN
D
+3V
3_G
PS
GN
D
TP19
TP18
SF11
86K
-3
INC
GN
DB
OU
TA
GN
DD
GN
DE
U24
+3V
3_G
PS
8.2p
F
C85
8.2p
F
C84
GN
D
VC
C_R
F
AN
T_O
N
RF_
OU
T2
VC
C_R
F RF_
OU
T2
10K
R65
GN
D
0R64
AN
T_O
N
8.2p
F
C86
RF_
OU
T1
0R66
0R63
0R62
RF_
OU
T1
GPS
Mod
ule
and
SAW
Filt
er
Low
Noi
se A
mpl
ifier
(LN
A)
Act
ive
Ant
enna
Sup
ply
RF_
LNA
0R67
RF_
LNA
10R68
LQW
15A
N27
NJ0
0D
L7 27nH
VC
C_R
FR
F_O
UT1
Ant
enna
Sel
ectio
n C
ircui
t
JP10
[1] A
nten
na tr
aces
requ
ire 5
0ohm
im
peda
nce
all t
he w
ay th
roug
h fr
om th
e G
PS m
odul
e to
the
ante
nna
conn
ecto
r
PIC8301PIC8302CO
C83
PIC8
401
PIC8
402COC
84
PIC8501
PIC8
502COC8
5PI
C860
1PI
C860
2COC8
6
PIC8701 PIC8702
COC8
7
PIC8801 PIC8802CO
C88
PIC8901
PIC8902COC
89
PID401 PID402COD4
PIJ1
701
PIJ170GND
COJ17
PIJ1
801
PIJ180GND
COJ18
PIJP1001PIJP1002CO
JP10
PIL601
PIL602
COL6
PIL701
PIL702
COL7
PIR6101 PIR6102COR6
1
PIR6
201
PIR6
202COR6
2
PIR6301PIR6302 COR6
3
PIR6
401
PIR6
402
COR6
4
PIR6501PIR6502 COR6
5
PIR6601PIR6602 COR6
6
PIR6
701
PIR6
702COR6
7
PIR6801
PIR6802
COR6
8
PITP18
01COTP18
PITP19
01COTP19
PIU2301
PIU2302
PIU2303
PIU2304
PIU2305
PIU2306
PIU2307
PIU2308
PIU2309
PIU23010
PIU23011
PIU23012
PIU23013
PIU23014
PIU23015
PIU23016
PIU23017
PIU23018CO
U23
PIU240A
PIU240B
PIU240C
PIU240D
PIU240ECO
U24
PIU2501
PIU2502
PIU2503
PIU2504
PIU2505
PIU2506
COU2
5
PIJP1002PIC8302
PIJP1001
PIR6101
PIU2307
PIU2308
PIU2309
PIR6
401
PIU23013
NLAN
T0ON
PIC8301
PIC8702 PIC8802
PIJ170GND PIJ180GND
PIR6501
PIU2301
PIU23010
PIU23012
PIU240B
PIU240D
PIU240E
PIU2501
PIU2502
PIU2303
NLGPS0RX
POGPS0RX
PIU2302
NLGPS0TX
POGPS0TX
PIC8
402
PIC8
602
PIU240C
PIC8501
PIU240A
PIC8
502
PIU23011
PIC8701PID402PIU2306
PIC8901
PIL601
PIC8902
PIU2503
PID401PIR6102
PIJ1
701
PIR6
201
PIJ1
801
PIR6
701
PIL701
PIR6802
PIR6
202
PIR6301 PIR6602
PIR6
702
PIR6
402
PIR6502PIU2505
PITP18
01PIU2305
PITP19
01PIU2304
PIU23015
PIU23016
PIU23017
PIU23018
PIL602
PIR6302
NLRF0LNA
PIC8
601
PIL702
PIR6601
NLRF0OUT1
PIC8
401
PIU2506
NLRF0OUT2
PIC8801
PIR6801
PIU23014
PIU2504
POGPS0RX
POGPS0TX
139
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
AC
AN
, JT
AG
, LD
O, U
SB.S
chD
oc2.
020
-DEC
-201
57
16En
g. J
orge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
2 4 6 8 10
1 3 5 7 9
JTA
G1
+3V
3
GN
D
JTA
G
TMS
TCK
TDI
TDO
RST
TXD
1
GN
D2
VC
C3
RX
D4
SHD
N5
CA
NL
6C
AN
H7
RS
8
MA
X30
51EK
A+T
U5
GN
D
120
R16
GN
D
CA
NH
CA
NL
+3V
3
GN
D
CA
N T
rans
ceiv
er
RS
LDO
Vol
tage
Reg
ulat
or (M
ain)
IN1
OU
T3
GN
D 24
LM39
40IM
PX-3
.3/N
OPB
U6
GN
DG
ND
GN
D
+3V
3+5
V
Gre
enLE
D6
GN
D
+3V
3
LDO
Vol
tage
Reg
ulat
or (S
enso
rs)
LP29
92A
IM5-
3.3/
NO
PB
IN1
GN
D2
ON
/OFF
3
BY
PASS
4
OU
T5
U7
GN
D
4.7u
FC
242.
2uF
C23
GN
D
+5V
GN
D
GN
D
+3V
3_SE
NS
100n
FC
20
0R17
470
R18
100n
FC
2233
uFC
21
TMS
TCK
TDO
TDI
RST
JTA
G
JTA
G
TX RXC
AN
CA
NC
AN
HC
AN
L
CA
N_B
US
CA
N_B
US
VB
US
1
D-
2
D+
3
ID4
GN
D5
SHSH
USB
1
Mic
ro U
SB B
GN
D
0D
NP
R20
GN
D
0D
NP
R21
+5V
0D
NP
R19
CG0603MLC-05LECT-ND
Z1
10K
R23
CG0603MLC-05LECT-ND
Z2
GN
DG
ND
GN
D
1N58
19H
W-F
DIC
T-N
DD
3
VB
US
D_N
D_PU
SB
D_N
USB
D_P
USB
_D_N
USB
_D_P
USB
100K
R22
10nF
C25
PIC2001 PIC2002CO
C20
PIC2101 PIC2102CO
C21
PIC2201 PIC2202CO
C22
PIC2301PIC2302CO
C23
PIC2401PIC2402COC
24
PIC2501 PIC2502CO
C25
PID301PID302COD3
PIJT
AG10
1PI
JTAG
102
PIJT
AG10
3PI
JTAG
104
PIJT
AG10
5PI
JTAG
106
PIJT
AG10
7PI
JTAG
108
PIJT
AG10
9PIJTAG1010
COJT
AG1
PILED601 PILED602CO
LED6
PIR1601PIR1602 COR1
6
PIR1701PIR1702 COR1
7
PIR1801 PIR1802COR1
8
PIR1
901
PIR1902
COR1
9PI
R200
1PIR2002
COR2
0
PIR2
101
PIR2
102
COR2
1
PIR2201
PIR2
202
COR2
2
PIR2301PIR2302 COR2
3
PIU501
PIU502
PIU503
PIU504
PIU505
PIU506
PIU507
PIU508
COU5
PIU601
PIU602
PIU603
PIU604
COU6
PIU701
PIU702
PIU703
PIU704
PIU705
COU7
PIUSB101
PIUSB102
PIUSB103
PIUSB104
PIUSB105
PIUSB10SH
COUS
B1
PIZ101PIZ102COZ
1PIZ201PIZ202
COZ2
PIC2001
PIC2101
PIJT
AG10
1
PIR1801PIU503
PIU603
PIC2402PIU705
PIC2201 PIC2302
PID302PIU601
PIU701
PIU703
PIR1602PIU507
NLCA
NH
POCAN0BUS
PIR1601PIU506
NLCA
NL
POCAN0BUS
PIR2
001
NLD0
NPOUSB
PIR2
101NLD0P
POUSB
PIC2002
PIC2102PIC2202 PIC2301
PIC2401
PIC2502
PIJT
AG10
3
PIJT
AG10
5
PIJT
AG10
7
PIJT
AG10
9
PILED602
PIR1701
PIR2201
PIR2301
PIU502
PIU602PIU604
PIU702
PIUSB105
PIUSB10SH
PIZ101PIZ201
PIC2501PIU704
PID301PI
R190
1
PIR2302
PIUSB101
PILED601PIR1802
PIR1902
POUSB
PIR2
202
PIUSB104
PIU501
POCAN
PIU504
POCAN
PIU505
PIR1702PIU508
NLRS
PIJTAG1010
NLRST
POJTAG
PIJT
AG10
4NL
TCK
POJTAG
PIJT
AG10
8NL
TDI
POJTAG
PIJT
AG10
6NL
TDO
POJTAG
PIJT
AG10
2NL
TMS
POJTAG
PIR2002
PIUSB102
PIZ102
NLUSB0D0N
PIR2
102
PIUSB103
PIZ202NL
USB0
D0P
POCAN
POCAN0RX
POCAN0TX
POCAN0BUS
POCAN0BUS0CANH
POCAN0BUS0CANL
POJTAG
POJTAG0RST
POJTAG0TCK
POJTAG0TDI
POJTAG0TDO
POJTAG0TMS
POUSB
POUSB0D0N
POUSB0D0P
POUSB0VBUS
140
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
ATe
lem
etry
, I2C
.Sch
Doc
2.0
20-D
EC-2
015
816
Eng.
Jor
ge M
arin
(mar
inm
ar@
ualb
erta
.ca)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
XB
P9B
-DM
ST-0
02
VC
C1
DO
UT
2
DIN
3
DO
84
RES
ET5
PWM
0/R
SSI
6
PWM
17
(Res
erve
d)8
DTR
/SLE
EP_R
Q/D
I89
GN
D10
AD
4/D
IO4
11C
TS/D
IO7
12O
N/S
LEEP
13V
REF
14A
SSO
C/A
D5/
DIO
515
RTS
/AD
6/D
IO6
16A
D3/
DIO
317
AD
2/D
IO2
18A
D1/
DIO
119
AD
0/D
IO0
20
U12
+3V
3
GN
D
RSS
I
XB
EE_R
XX
BEE
_TX
Tele
met
ry (X
Bee)
I2C
1 Le
vel S
hifte
r
PCA
9306
DC
TR
GN
D1
VR
EF1
2
SCL1
3
SDA
14
SDA
25
SCL2
6V
REF
27
EN8
U13
GN
D
81 2.2K
RN
4A
+3V
3+3
V3
+3V
3+5
V+5
V+5
V
200K
R38
I2C
2 Le
vel S
hifte
r
GN
D
+3V
3+3
V3
+3V
3+5
V+5
V+5
V
PCA
9306
DC
TR
GN
D1
VR
EF1
2
SCL1
3
SDA
14
SDA
25
SCL2
6V
REF
27
EN8
U14
200K
R39
72 2.2K
RN
4B
63 2.2K
RN
4C
54 2.2K
RN
4D
63 2.2K
RN
5C
54 2.2K
RN
5D
81 2.2K
RN
5A
72 2.2K
RN
5B
0R
400
R41
UA
RT
(Low
Vol
tage
)6
2
1
BSS
138P
S,11
5Q
2A
81 10K
RN
3A
72 10K
RN
3B
+3V
3LV
LV_T
X
3
5
4
BSS
138P
S,11
5Q
2B
63 10K
RN
3C
54 10K
RN
3D
+3V
3LV
LV_R
X
SDA
SCL
I2C
1
I2C
1
SDA
SCL
I2C
2
I2C
2
SDA
SCL
I2C
1_5V
I2C
1_5V
SDA
SCL
I2C
2_5V
I2C
2_5V
UA
RT_
TX
UA
RT_
RX
1 2 3 4 5 6 7 8 9 10
EX1
Hea
der 1
0 2m
m
XB
EE_T
X
1 2 3 4 5 6 7 8 9 10
EX2
Hea
der 1
0 2m
m
XB
EE_R
X
GN
D
+3V
3
XB
EE_T
XX
BEE
_RX
GN
DR
SSI
Yel
low
LED
7
PIN
5PI
N4
PIN
7PI
N8
PIN
9
RSS
IPI
N20
PIN
11PI
N12
PIN
13PI
N14
PIN
15PI
N16
PIN
17PI
N18
PIN
19
PIN
20
PIN
11PI
N12
PIN
13PI
N14
PIN
15PI
N16
PIN
17PI
N18
PIN
19
PIN
5PI
N4
PIN
7PI
N8
PIN
9
63
470
RN
6C
PIEX101
PIEX102
PIEX103
PIEX104
PIEX105
PIEX106
PIEX107
PIEX108
PIEX109
PIEX1010
COEX1
PIEX201
PIEX202
PIEX203
PIEX204
PIEX205
PIEX206
PIEX207
PIEX208
PIEX209
PIEX2010
COEX2
PILED701
PILED702
COLE
D7
PIQ201
PIQ202
PIQ206
COQ2A
PIQ203
PIQ204
PIQ205
COQ2B
PIR3801 PIR3802COR3
8
PIR3901 PIR3902COR3
9
PIR4
001
PIR4
002
COR4
0PI
R410
1PI
R410
2CO
R41
PIRN301 PIRN308CORN3A
PIRN302 PIRN307
CORN3B
PIRN303 PIRN306CORN3C
PIRN304 PIRN305
CORN3D
PIRN401 PIRN408CORN4A
PIRN402 PIRN407
CORN4B
PIRN403 PIRN406
CORN4C
PIRN404 PIRN405
CORN4D
PIRN501 PIRN508CORN5A
PIRN502 PIRN507
CORN5B
PIRN503 PIRN506
CORN5C
PIRN504 PIRN505
CORN5D
PIRN603
PIRN
606CORN6C
PIU1201
PIU1202
PIU1203
PIU1204
PIU1205
PIU1206
PIU1207
PIU1208
PIU1209
PIU12010
PIU12011
PIU12012
PIU12013
PIU12014
PIU12015
PIU12016
PIU12017
PIU12018
PIU12019
PIU12020
COU12
PIU1301
PIU1302
PIU1303
PIU1304
PIU1305
PIU1306
PIU1307
PIU1308
COU1
3
PIU1401
PIU1402
PIU1403
PIU1404
PIU1405
PIU1406
PIU1407
PIU1408
COU14
PIEX101
PIRN302PIRN304
PIRN401PIRN402
PIRN501PIRN502
PIU1201
PIU1302
PIU1402
PIR3801 PIR3901
PIRN403PIRN404
PIRN503PIRN504
PIEX1010
PILED702
PIU12010
PIU1301
PIU1401
PIQ202 PIQ205
PIRN301 PIRN303
PIQ204
PIRN306POLV0RX
PIQ201
PIRN308POLV0TX
PILED701
PIRN603
PIR3802PIU1307
PIU1308
PIR3902PIU1407
PIU1408
PIR4
001
PIU1403
PIR4
002
PIRN507
POI2C2
PIR4
101PIU1404
PIR4
102
PIRN508
POI2C2
PIRN405
PIU1305
POI2C105V
PIRN406
PIU1306
POI2C105V
PIRN407
PIU1303
POI2C1
PIRN408
PIU1304
POI2C1
PIRN505
PIU1405
POI2C205V
PIRN506
PIU1406
POI2C205V
PIEX104
PIU1204
NLPIN4
PIEX105
PIU1205
NLPIN5
PIEX107
PIU1207
NLPIN7
PIEX108
PIU1208
NLPIN8
PIEX109
PIU1209
NLPIN9
PIEX2010
PIU12011NL
PIN1
1
PIEX209
PIU12012NL
PIN1
2
PIEX208
PIU12013NL
PIN1
3
PIEX207
PIU12014NL
PIN1
4
PIEX206
PIU12015NL
PIN1
5
PIEX205
PIU12016NL
PIN1
6
PIEX204
PIU12017NL
PIN1
7
PIEX203
PIU12018NL
PIN1
8
PIEX202
PIU12019NL
PIN1
9
PIEX201
PIU12020NL
PIN2
0PIEX106
PIRN
606
PIU1206
NLRSSI
PIQ203
PIRN307POUART0RX
PIQ206
PIRN305POUART0TX
PIEX103
PIU1203
NLXBEE0RX
POXBEE0RX
PIEX102
PIU1202
NLXBEE0TX
POXBEE0TX
POI2C1
POI2C10SCL
POI2C10SDA
POI2C105V
POI2
C105
V0SC
LPO
I2C1
05V0
SDA
POI2C2
POI2C20SCL
POI2C20SDA
POI2C205V
POI2
C205
V0SC
LPO
I2C2
05V0
SDA
POLV0RX
POLV0TX
POUART0RX
POUART0TX
POXBEE0RX
POXBEE0TX
141
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
ASw
itche
s, U
I.Sc
hDoc
2.0
20-D
EC-2
015
916
Eng.
Jor
ge M
arin
(mar
inm
ar@
ualb
erta
.ca)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
BU
FF1_
EN
Serv
o/M
otor
Sig
nals
Ena
ble
Switc
hes
Red
LED
1O
rang
eLE
D2
Yel
low
LED
4
+3V
3+3
V3
+3V
3+3
V3
+3V
3
LED
s
Gre
enLE
D3
Red
LED
5 GN
D
+3V
3
BIN
D
Bin
d B
utto
nS1 BIN
D
10K
R44
VC
C14
A1
2
GN
D7
OE1
1
Y1
3
A2
5
Y4
11
OE2
4
OE4
13
Y2
6
A4
12A
39
Y3
8
OE3
10
U15
SN74
LVC
125A
PWR
10K
R42
+3V
3
GN
D
S1S1 S2 S3 S4
S2 S3 S4
+3V
3
GN
D
S5S5 S6 S7 S8
S6 S7 S8
S1_E
N
VC
C14
A1
2
GN
D7
OE1
1
Y1
3
A2
5
Y4
11
OE2
4
OE4
13
Y2
6
A4
12A
39
Y3
8
OE3
10
U16
SN74
LVC
125A
PWR
LED
s
LED
1LE
D2
LED
3LE
D4
LED
5
LED
s
Serv
o1Se
rvo2
Serv
o3Se
rvo4
Serv
o5Se
rvo6
Serv
o7Se
rvo8
Serv
os
Serv
os
S2_E
NS3
_EN
S4_E
N
S5_E
NS6
_EN
S7_E
NS8
_EN
3 1
2SB6 S1_EN
3 1
2SB7
3 1
2SB8
3 1
2SB9
3 1
2SB10
3 1
2SB11
3 1
2SB12
S2_EN
S3_EN
S4_EN
S5_EN
S6_EN
S7_EN
S8_EN
3 1
2SB13
BU
FF2_
EN
GN
D
10K
R43
GN
D
5 4470
RN
7D
6 3470
RN
7C
7 2470
RN
7B
8 1470
RN
7A
5 4470
RN
6D
81
1KR
N8A
S1
72
1KR
N8B
S2
63
1KR
N8C
S3
54
1KR
N8D
S4G
ND
81
1KR
N9A
S5
72
1KR
N9B
S6
63
1KR
N9C
S7
54
1KR
N9D
S8G
ND
PILED101 PILED102CO
LED1
PILED201 PILED202CO
LED2
PILED301 PILED302CO
LED3
PILED401 PILED402COLED4
PILED501 PILED502CO
LED5
PIR4201
PIR4
202COR4
2
PIR4301
PIR4
302COR4
3
PIR4401PIR4402 COR4
4
PIRN604PIRN605 CORN6D
PIRN701PIRN708 CORN7A
PIRN702PIRN707CORN7B
PIRN703PIRN706CORN7C
PIRN704PIRN705CORN7D
PIRN801
PIRN
808
CORN8A
PIRN802
PIRN
807
CORN8B
PIRN803
PIRN
806
CORN8C
PIRN804
PIRN
805
CORN8D
PIRN901
PIRN908
CORN9A
PIRN902
PIRN907
CORN9B
PIRN903
PIRN906
CORN9C
PIRN904
PIRN905
CORN9D
PIS101PIS102 COS1
PISB601PISB602
PISB603COSB6
PISB701PISB
702
PISB703COSB7
PISB801PISB
802
PISB803COSB8
PISB901PISB
902
PISB903CO
SB9
PISB1001PISB10
02
PISB1003CO
SB10
PISB1101PISB1102
PISB1103CO
SB11
PISB1201PISB1202
PISB1203CO
SB12
PISB1301PISB13
02
PISB1303CO
SB13
PIU1501
PIU1502
PIU1503
PIU1504
PIU1505
PIU1506
PIU1507
PIU1508
PIU1509
PIU15010
PIU15011
PIU15012
PIU15013
PIU15014
COU1
5
PIU1601
PIU1602
PIU1603
PIU1604
PIU1605
PIU1606
PIU1607
PIU1608
PIU1609
PIU16010
PIU16011
PIU16012
PIU16013
PIU16014
COU1
6PIR4402
PIRN605PIRN705
PIRN706PIRN707
PIRN708
PIU15014
PIU16014
PIR4401 PIS102POBIND
PIR4301
PISB601PISB701
PISB801PISB901
PISB1001PISB1101
PISB1201PISB1301
POBUFF10ENPI
R4201
PISB603PISB703
PISB803PISB903
PISB1003PISB1103
PISB1203PISB1303
POBU
FF20
EN
PIR4
202
PIR4
302
PIRN
805
PIRN
806
PIRN
807
PIRN
808
PIRN905
PIRN906
PIRN907
PIRN908
PIS101
PIU1507
PIU1607
PILED101PIRN704 PILED102
POLEDs
PILED201PIRN703 PILED202
POLEDs
PILED301PIRN702 PILED302
POLEDs
PILED401PIRN701 PILED402
POLEDs
PILED501PIRN604 PILED502
POLEDs
PIU1502
POServos
PIU1505
POServos
PIU1509
POServos
PIU15012
POServos
PIU1602
POServos
PIU1605
POServos
PIU1609
POServos
PIU16012
POServos
PIRN801
PIU1503
NLS1
POS1
PISB602
PIU1501
NLS1
0EN
PIRN802
PIU1506
NLS2
POS2
PISB
702
PIU1504
NLS2
0EN
PIRN803
PIU1508
NLS3
POS3
PISB
802
PIU15010
NLS3
0EN
PIRN804
PIU15011
NLS4
POS4
PISB
902
PIU15013
NLS40EN
PIRN901
PIU1603
NLS5
POS5
PISB10
02
PIU1601
NLS50EN
PIRN902
PIU1606
NLS6
POS6
PISB1102
PIU1604
NLS60EN
PIRN903
PIU1608
NLS7
POS7
PISB1202
PIU16010
NLS70EN
PIRN904
PIU16011
NLS8
POS8
PISB13
02
PIU16013
NLS80EN
POBIND
POBUFF10EN
POBU
FF20
EN
POLEDS
POLE
DS0L
ED1
POLE
DS0L
ED2
POLE
DS0L
ED3
POLE
DS0L
ED4
POLE
DS0L
ED5
POS1
POS2
POS3
POS4
POS5
POS6
POS7
POS8
POSERVOS
POSERV
OS0SER
VO1POS
ERVOS0
SERVO2
POSERV
OS0SER
VO3POS
ERVOS0
SERVO4
POSERV
OS0SER
VO5POS
ERVOS0
SERVO6
POSERV
OS0SER
VO7POS
ERVOS0
SERVO8
142
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
AD
ata
Logg
er.S
chD
oc2.
020
-DEC
-201
510
16En
g. J
orge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
D2
- N
C1
CD/D
3- C
S2
CD
9
GND 10
CMD
- D
I3
VDD
4
CLK
- S
CK5
VSS
6
D0
- D
O7
D1
- N
C8
MIC
RO
SD
U17
GN
D
SPI_
SCK
SPI_
MIS
O
SPI_
MO
SISP
I_C
S
GN
D
+3V
3_D
LC
D GN
D
+3V
3_D
L
Mic
ro S
D C
ard
1 2 3 4
J15
ICSP 1 2 3 4 5 6
J16
FTD
I
SPI_
SCK
SPI_
MIS
OSP
I_M
OSI
RES
ET
DTR
GN
DG
ND
+3V
3_D
L
TXO
RX
I
Hea
ders
PC6
(RES
ET/P
CIN
T14)
29
PD0
(RX
D/P
CIN
T16)
30
PD1
(TX
D/P
CIN
T17)
31
PD2
(IN
T0/P
CIN
T18)
32
PD4
(PC
INT2
0/X
CK
/T0)
2
VC
C6
GN
D5
PB6
(PC
INT6
/XTA
L1/T
OSC
1)7
PB7
(PC
INT7
/XTA
L2/T
OSC
2)8
PD5
(PC
INT2
1/O
C0B
/T1)
9
PD6
(PC
INT2
2/O
C0A
/AIN
0)10
PD7
(PC
INT2
3/A
IN1)
11
PB0
(PC
INT0
/CLK
O/IC
P1)
12
PB1
(PC
INT1
/OC
1A)
13
PB2
(PC
INT2
/SS/
OC
1B)
14
PB3
(PC
INT3
/OC
2A/M
OSI
)15
PB4
(PC
INT4
/MIS
O)
16
PB5
(SC
K/P
CIN
T5)
17
AV
CC
18
AR
EF20
GN
D21
PC0
(AD
C0/
PCIN
T8)
23
PC1
(AD
C1/
PCIN
T9)
24
PC2
(AD
C2/
PCIN
T10)
25
PC3
(AD
C3/
PCIN
T11)
26
PC4
(AD
C4/
SDA
/PC
INT1
2)27
PC5
(AD
C5/
SCL/
PCIN
T13)
28
GN
D3
VC
C4
AD
C6
19
AD
C7
22
PD3
(PC
INT1
9/O
C2B
/INT1
)1
U18
ATm
ega3
28P-
AU
GN
D
GN
D
SPI_
SCK
SPI_
MIS
OSP
I_M
OSI
SPI_
CS
RES
ET
CD
TXO
RX
I
Dat
a Lo
gger
MC
U
DTR
+3V
3_D
L
+3V
3_D
L
+3V
3_D
L
GN
DG
ND
RX
IU
AR
T_IN
100n
FC
54
100n
F
C52
10K
R45
12
16M
Hz
Y2
GN
DG
reen
LED
8G
reen
LED
9
0R46
+3V
3_D
L+3
V3
7 2470
RN
6B
8 1470
RN
6A
JP7
100n
FC
5510
0nF
C56
100n
FC
53
LED
10 to
ggle
s on
/off
ev
ery
time
a ne
w
char
acte
r is
rece
ived
LED
9 is
con
nect
ed to
th
e SP
I SC
K li
ne. I
t fla
shes
ver
y fa
st w
hen
writ
ing
to th
e SD
car
d
PIC5201
PIC5202
COC5
2PIC5301 PIC5302
COC5
3
PIC5401 PIC5402CO
C54
PIC5501 PIC5502CO
C55
PIC5601 PIC5602COC56
PIJ1
501
PIJ1
502
PIJ1
503
PIJ1
504COJ
15
PIJ1601
PIJ1
602
PIJ1603
PIJ1
604
PIJ1
605
PIJ1
606COJ
16
PIJP
701
PIJP
702
COJP7
PILED801 PILED802CO
LED8
PILED901 PILED902CO
LED9
PIR4501PIR4502 COR4
5
PIR4
601
PIR4
602COR4
6
PIRN601PIRN608 CORN6A
PIRN602PIRN607CORN6B
PIU1701
PIU1702
PIU1703
PIU1704
PIU1705
PIU1706
PIU1707
PIU1708
PIU1709
PIU17010
COU17
PIU1801
PIU1802
PIU1803
PIU1804
PIU1805
PIU1806
PIU1807
PIU1808
PIU1809
PIU18010
PIU18011
PIU18012
PIU18013
PIU18014
PIU18015
PIU18016
PIU18017
PIU18018
PIU18019
PIU18020
PIU18021
PIU18022
PIU18023
PIU18024
PIU18025
PIU18026
PIU18027
PIU18028
PIU18029
PIU18030
PIU18031
PIU18032
COU18
PIY2
01PIY202
PIY203
COY2
PIJP
702
PIC5301
PIC5401PIC5501
PIC5601
PIJ1603
PIJP
701
PIR4502PIU1704
PIU1804
PIU1806
PIU18018
PIU18020
NL03
V30D
L
PIU1709
PIU18025
NLCD
PIC5202
PIJ1
606
NLDT
R
PIC5302
PIC5402PIC5502
PIC5602PIJ1601
PIJ1
602
PILED802PILED902
PIU1706
PIU17010
PIU1803
PIU1805
PIU18021
PIY203
NLGND
PILED801PIRN601
PILED901PIRN602PIRN607
PIU1809
PIU1701
PIU1708
PIU1801
PIU1802
PIU1807
PIY2
01
PIU1808
PIY202
PIU18010
PIU18011
PIU18012
PIU18013
PIU18019
PIU18022
PIU18023
PIU18024
PIU18026
PIU18027
PIU18028
PIU18032
PIC5201
PIJ1
501
PIR4501
PIU18029
NLRE
SET
PIJ1
604
PIR4
602
PIU18030
NLRXI
PIU1702
PIU18014
NLSP
I0CS
PIJ1
504
PIU1707
PIU18016
NLSPI0MISO
PIJ1
503
PIU1703
PIU18015
NLSPI0MOSI
PIJ1
502
PIRN608
PIU1705
PIU18017
NLSP
I0SC
K
PIJ1
605
PIU18031
NLTX
O
PIR4
601
POUART0IN
POUART0IN
143
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
APo
wer
.Sch
Doc
2.0
20-D
EC-2
015
1116
Eng.
Jor
ge M
arin
(mar
inm
ar@
ualb
erta
.ca)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
DC
1_EN
DC
2_EN
U_M
ain
Pow
erM
ain
Pow
er.S
chD
oc
VIN
GN
D
VIN
_BC
KP
DC
3_EN
DC
4_EN
U_B
acku
p Po
wer
Bac
kup
Pow
er.S
chD
ocB
AC
KU
P_EN
+5V
_BC
KP
+5V
_SER
VO
_BC
KP
+5V
_MA
IN
+5V
_SER
VO
CM
S04(
TE12
L,Q
,M)
D1
CM
S04(
TE12
L,Q
,M)
D2
1 2 3 4
J1 POW
ER
VIN
GN
D
+5V
_BC
KP
+5V
_SER
VO
_BC
KP
12
34
56
78
J4 Hea
der 4
X2
+5V
_SER
VO
+5V
_MA
IN
GN
DV
IN
DC
1_EN
DC
2_EN
BA
CK
UP_
EN
VIN
+5V
_BC
KP
+5V
_SER
VO
_BC
KP
BA
CK
UP_
EN
VIN
GN
D
+5V
_BC
KP
+5V
_SER
VO
_BC
KP
VIN
+5V
_BC
KP
+5V
_SER
VO
_BC
KP
BA
CK
UP_
EN
Mai
n Po
wer
Bac
kup
Pow
er
12
34
56
78
J2 Hea
der 4
X2
12
34
56
78
J8 Hea
der 4
X2
DC
1_EN
DC
2_EN
+5V
_EX
T
PID101
PID102
COD1
PID201
PID202
COD2
PIJ101
PIJ102
PIJ103
PIJ104
COJ1
PIJ201
PIJ202
PIJ203
PIJ204
PIJ205
PIJ206
PIJ207
PIJ208
COJ2
PIJ401
PIJ402
PIJ403
PIJ404
PIJ405
PIJ406
PIJ407
PIJ408
COJ4
PIJ801
PIJ802
PIJ803
PIJ804
PIJ805
PIJ806
PIJ807
PIJ808
COJ8
PID101
PIJ407
PIJ408
PIJ807
PIJ808NL05V0BCKP
PIJ205
PID102
PIJ207
NL05V0MAIN
PID202
PIJ203
NL05V0SERVO
PID201
PIJ405
PIJ406
PIJ805
PIJ806NL05V0SERVO0BCKP
PIJ204
PIJ403
PIJ804
NLBA
CKUP
0EN
PIJ208
NLDC10EN
PIJ206
NLDC20EN
PIJ102
PIJ104
PIJ201
PIJ404
PIJ803
NLGN
D
PIJ101
PIJ202
PIJ401
PIJ402
PIJ801
PIJ802NL
VIN
PIJ103
144
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
AM
ain
Pow
er.S
chD
oc2.
020
-DEC
-201
512
16En
g. J
orge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
SW1
SW2
CB
OO
T3
VC
C4
BIA
S5
SYN
C6
RT
7
PGO
OD
8
FB9
AG
ND
10
SS/T
RK
11
ENA
BLE
12
VIN
13
VIN
14
PGN
D15
PGN
D16
PAD
U8
LM43
603P
WPT
GN
DG
ND
DN
PR
29
GN
D
10uF
C28
453K
R25
100K
R30
GN
D
4.7u
FC
29
GN
DG
ND
GN
D
VIN
TP11
0.47
uF
C27
10nF
C37
GN
D
IHLP
3232
DZE
R6R
8M11
L2 6.8u
H
4.7u
FC
38
GN
D
1Meg
R24
100p
FC
30
249K
R28
GN
D
100K
R27
47uF
C31
10uF
C33
100n
FC
34
GN
DG
ND
GN
DG
ND
DN
PC
26
+5V
_MA
IN
Mai
n Po
wer
- Lo
cal S
tep-
Dow
n V
olta
ge C
onve
rter
Mai
n Po
wer
- E
xter
nal S
tep-
Dow
n V
olta
ge C
onve
rter
+5V
_EX
T
2.2u
FC
36
0R26
47uF
C32
SW1
SW2
CB
OO
T3
VC
C4
BIA
S5
SYN
C6
RT
7
PGO
OD
8
FB9
AG
ND
10
SS/T
RK
11
ENA
BLE
12
VIN
13
VIN
14
PGN
D15
PGN
D16
PAD
U10
LM43
603P
WPT
GN
DG
ND
GN
D
10uF
C41
453K
R32
100K
R37
GN
D
4.7u
FC
42
GN
DG
ND
GN
D
VIN
TP13
0.47
uF
C40
10nF
C50
GN
D
IHLP
3232
DZE
R6R
8M11
L3 6.8u
H
4.7u
FC
51
GN
D
1Meg
R31
249K
R35
GN
D
100K
R34
47uF
C44
10uF
C46
100n
FC
47
GN
DG
ND
GN
DG
ND
2.2u
FC
49
47uF
C45
DN
PC
39
DN
PR
36
0R33
VO
UT_
5V_M
AIN
_LO
CA
L
VO
UT_
5V_M
AIN
_EX
T
DC
1_EN
2.2V
Thr
esho
ld
2.2V
Thr
esho
ld DC
2_EN
ENA
BLE
1
VIN
2
GN
D3
GN
D4
VO
UT
5
D24
V50
FX
U9
DC
1_EN
DC
2_EN
DC
1_EN
VIN
GN
D
DC
1M_V
OU
T
DC
1M_V
OU
T
DC
2M_V
OU
T
ENA
BLE
1
VIN
2
GN
D3
GN
D4
VO
UT
5
D24
V50
FX
U11
DC
2_EN
VIN
GN
D
DC
2M_V
OU
T
TP10
TP12
JP5
JP6
100n
FC
35
100n
FC
4810
0pF
C43
R25
and
R30
form
a v
olta
ge d
ivid
er u
sed
for c
utof
f vo
ltage
. If t
he v
olta
ge o
n th
e EN
AB
LE p
in is
low
er th
an
2.2V
the
LM43
603
will
shu
t dow
n. R
25 c
an b
e ch
ange
d to
a
high
er v
alue
to in
crea
se th
e cu
toff
vol
tage
.
R32
and
R37
form
a v
olta
ge d
ivid
er u
sed
for c
utof
f vo
ltage
. If t
he v
olta
ge o
n th
e EN
AB
LE p
in is
low
er th
an
2.2V
the
LM43
603
will
shu
t dow
n. R
32 c
an b
e ch
ange
d to
a
high
er v
alue
to in
crea
se th
e cu
toff
vol
tage
.
The
jum
per i
s us
ed to
se
lect
bet
wee
n th
e on
bo
ard
regu
lato
r and
an
exte
rnal
regu
lato
r
The
jum
per i
s us
ed to
se
lect
bet
wee
n th
e on
bo
ard
regu
lato
r and
an
exte
rnal
regu
lato
r
PIC2
601
PIC2
602
COC2
6
PIC2
701
PIC2
702
COC2
7
PIC2801 PIC2802CO
C28
PIC2901 PIC2902CO
C29
PIC3001 PIC3002CO
C30
PIC3101 PIC3102CO
C31
PIC3201 PIC3202COC32
PIC3301 PIC3302CO
C33
PIC3401 PIC3402CO
C34
PIC3501 PIC3502CO
C35
PIC3601PIC3602CO
C36
PIC3701 PIC3702CO
C37
PIC3801 PIC3802CO
C38
PIC3
901
PIC3
902
COC3
9
PIC4
001
PIC4
002
COC40
PIC4101 PIC4102CO
C41
PIC4201 PIC4202CO
C42
PIC4301 PIC4302CO
C43
PIC4401 PIC4402CO
C44
PIC4501 PIC4502COC45
PIC4601 PIC4602CO
C46
PIC4701 PIC4702CO
C47
PIC4801 PIC4802CO
C48
PIC4901PIC4902CO
C49
PIC5001 PIC5002CO
C50
PIC5101 PIC5102CO
C51
PIJP501 PIJP
502
PIJP503
COJP5
PIJP601 PIJP
602
PIJP603
COJP6
PIL201
PIL202
COL2
PIL301
PIL302
COL3
PIR2401 PIR2402COR2
4
PIR2501PIR2502 COR2
5
PIR2
601
PIR2
602COR2
6
PIR2
701
PIR2
702COR2
7PIR2801 PIR2802CO
R28
PIR2901PIR2902 COR2
9
PIR3001PIR3002 COR3
0
PIR3101 PIR3102COR3
1
PIR3201PIR3202 COR3
2
PIR3
301
PIR3
302COR3
3
PIR3
401
PIR3
402COR3
4PIR3501 PIR3502CO
R35
PIR3601PIR3602 COR3
6
PIR3701PIR3702 COR3
7
PITP1001 COTP
10PITP1101
COTP11
PITP1201 COTP
12PITP1301
COTP13
PIU801
PIU802
PIU803
PIU804
PIU805
PIU806
PIU807
PIU808
PIU809
PIU8010
PIU8011
PIU8012
PIU8013
PIU8014
PIU8015
PIU8016
PIU80EP
COU8
PIU901
PIU902
PIU903
PIU904
PIU905CO
U9
PIU1001
PIU1002
PIU1003
PIU1004
PIU1005
PIU1006
PIU1007
PIU1008
PIU1009
PIU10010
PIU10011
PIU10012
PIU10013
PIU10014
PIU10015
PIU10016
PIU100EPCO
U10
PIU1101
PIU1102
PIU1103
PIU1104
PIU1105CO
U11
PIJP
602
PIJP
502
PIR2501 PIR3002PITP1101
PIU8012
PIU901
NLDC10EN
PODC10EN
PIJP503
PIU905
NLDC1M0VOUT
PIR3201 PIR3702PITP1301
PIU10012
PIU1101
NLDC20EN
PODC20EN
PIJP603
PIU1105
NLDC2M0VOUT
PIC2802PIC2902
PIC3102PIC3202
PIC3302PIC3402
PIC3502
PIC3601
PIC3702PIC3802
PIC4102PIC4202
PIC4402PIC4502
PIC4602PIC4702
PIC4802
PIC4901
PIC5002PIC5102
PIR2802PIR2901
PIR3001
PIR3502PIR3601
PIR3701
PIU806
PIU8010
PIU8015
PIU8016
PIU80EP
PIU903
PIU904
PIU1006
PIU10010
PIU10015
PIU10016
PIU100EP
PIU1103
PIU1104
NLGND
PIC2
601
PIC2
702
PIL201
PIU801
PIU802
PIC2
701
PIU803
PIC3002PIR2402 PIR2801
PITP1001PIU809
PIC3602
PIU804
PIC3701PIU8011
PIC3801
PIR2
602
PIU805
PIC3
901
PIC4
002
PIL301
PIU1001
PIU1002
PIC4
001
PIU1003
PIC4302PIR3102 PIR3501
PITP1201PIU1009
PIC4902
PIU1004
PIC5001PIU10011
PIC5101
PIR3
302
PIU1005
PIR2
702
PIU808
PIR2902
PIU807
PIR3
402
PIU1008
PIR3602
PIU1007
PIC2801PIC2901
PIC3501
PIC4101PIC4201
PIC4801
PIR2502 PIR3202
PIU8013
PIU8014
PIU902
PIU10013
PIU10014
PIU1102
NLVIN
PIC3
902
PIC4301PIC4401
PIC4501PIC4601
PIC4701
PIJP601
PIL302
PIR3101
PIR3
301
PIR3
401
NLVO
UT05
V0MA
IN0E
XT
PIC2
602
PIC3001PIC3101
PIC3201PIC3301
PIC3401
PIJP501
PIL202
PIR2401
PIR2
601
PIR2
701
NLVO
UT05
V0MA
IN0L
OCAL
PODC10EN
PODC20EN
145
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
ABa
ckup
Pow
er.S
chD
oc2.
020
-DEC
-201
513
16En
g. J
orge
Mar
in (m
arin
mar
@ua
lber
ta.c
a)
UA
AR
G A
utop
ilot.P
rjPc
bU
nive
rsity
of A
lber
ta A
eria
l Rob
otic
s Gro
up
SW1
SW2
CB
OO
T3
VC
C4
BIA
S5
SYN
C6
RT
7
PGO
OD
8
FB9
AG
ND
10
SS/T
RK
11
ENA
BLE
12
VIN
13
VIN
14
PGN
D15
PGN
D16
PAD
U19
LM43
603P
WPT
GN
DG
ND
DN
PR
52
GN
D
10uF
C59
453K
R48
100K
R53
GN
D
4.7u
FC
60
GN
DG
ND
GN
D
TP15
10nF
C68
GN
D
IHLP
3232
DZE
R6R
8M11
L4 6.8u
H
4.7u
FC
69
GN
D
1Meg
R47
249K
R51
GN
D
100K
R50
47uF
C62
10uF
C64
100n
FC
65
GN
DG
ND
GN
DG
ND
+5V
_BC
KP
Bac
kup
Pow
er -
Loca
l Ste
p-D
own
Vol
tage
Con
verte
r
Bac
kup
Pow
er -
Bac
kup
Serv
o St
ep-D
own
Vol
tage
Con
vert
er
+5V
_SER
VO
_BC
KP
2.2u
FC
67
0R49
47uF
C63
SW1
SW2
CB
OO
T3
VC
C4
BIA
S5
SYN
C6
RT
7
PGO
OD
8
FB9
AG
ND
10
SS/T
RK
11
ENA
BLE
12
VIN
13
VIN
14
PGN
D15
PGN
D16
PAD
U21
LM43
603P
WPT
GN
DG
ND
GN
D
10uF
C72
453K
R55
100K
R60
GN
D
4.7u
FC
73
GN
DG
ND
GN
D
TP17
0.47
uF
C71
10nF
C81
GN
D
IHLP
3232
DZE
R6R
8M11
L5 6.8u
H
4.7u
FC
82
GN
D
1Meg
R54
249K
R58
GN
D
100K
R57
47uF
C75
10uF
C77
100n
FC
78
GN
DG
ND
GN
DG
ND
2.2u
FC
80
47uF
C76
DN
PC
70
DN
PR
59
0R56
VO
UT_
5V_B
AC
KU
P_LO
CA
L
VO
UT_
5V_B
AC
KU
P_SE
RV
O
DC
3_EN
2.2V
Thr
esho
ld
2.2V
Thr
esho
ld DC
4_EN
VIN
_BC
KP
VIN
_BC
KP
ENA
BLE
1
VIN
2
GN
D3
GN
D4
VO
UT
5
D24
V50
FX
U20
DC
3_EN
VIN
_BC
KP
GN
D
DC
3M_V
OU
T
DC
3M_V
OU
T
DC
4M_V
OU
T
ENA
BLE
1
VIN
2
GN
D3
GN
D4
VO
UT
5
D24
V50
FX
U22
DC
4_EN
VIN
_BC
KP
GN
D
DC
4M_V
OU
T
DC
4_EN
DC
3_EN
TP14
TP16
JP8
JP9
100n
FC
66
100n
FC
79
100p
FC
61
100p
FC
74
DN
PC
57
0.47
uF
C58
R48
and
R53
form
a v
olta
ge d
ivid
er u
sed
for c
utof
f vo
ltage
. If t
he v
olta
ge o
n th
e EN
AB
LE p
in is
low
er th
an
2.2V
the
LM43
603
will
shu
t dow
n. R
48 c
an b
e ch
ange
d to
a
high
er v
alue
to in
crea
se th
e cu
toff
vol
tage
.
R55
and
R60
form
a v
olta
ge d
ivid
er u
sed
for c
utof
f vo
ltage
. If t
he v
olta
ge o
n th
e EN
AB
LE p
in is
low
er th
an
2.2V
the
LM43
603
will
shu
t dow
n. R
55 c
an b
e ch
ange
d to
a
high
er v
alue
to in
crea
se th
e cu
toff
vol
tage
.
The
jum
per i
s us
ed to
se
lect
bet
wee
n th
e on
bo
ard
regu
lato
r and
an
exte
rnal
regu
lato
r
The
jum
per i
s us
ed to
se
lect
bet
wee
n th
e on
bo
ard
regu
lato
r and
an
exte
rnal
regu
lato
r
PIC5
701
PIC5
702
COC5
7
PIC5
801
PIC5
802
COC5
8
PIC5901 PIC5902CO
C59
PIC6001 PIC6002CO
C60
PIC6101 PIC6102CO
C61
PIC6201 PIC6202CO
C62
PIC6301 PIC6302COC63
PIC6401 PIC6402CO
C64
PIC6501 PIC6502CO
C65
PIC6601 PIC6602CO
C66
PIC6701PIC6702CO
C67
PIC6801 PIC6802CO
C68
PIC6901 PIC6902CO
C69
PIC7
001
PIC7
002
COC7
0
PIC7
101
PIC7
102
COC71
PIC7201 PIC7202CO
C72
PIC7301 PIC7302CO
C73
PIC7401 PIC7402CO
C74
PIC7501 PIC7502CO
C75
PIC7601 PIC7602COC76
PIC7701 PIC7702CO
C77
PIC7801 PIC7802CO
C78
PIC7901 PIC7902CO
C79
PIC8001PIC8002CO
C80
PIC8101 PIC8102CO
C81
PIC8201 PIC8202CO
C82
PIJP801 PIJP
802
PIJP803
COJP8
PIJP901 PIJP
902
PIJP903
COJP9
PIL401
PIL402
COL4
PIL501
PIL502
COL5
PIR4701 PIR4702COR4
7
PIR4801PIR4802 COR4
8
PIR4
901
PIR4
902COR4
9
PIR5
001
PIR5
002COR5
0PIR5101 PIR5102CO
R51
PIR5201PIR5202 COR5
2
PIR5301PIR5302 COR5
3
PIR5401 PIR5402COR5
4
PIR5501PIR5502 COR5
5
PIR5
601
PIR5
602COR5
6
PIR5
701
PIR5
702COR5
7PIR5801 PIR5802CO
R58
PIR5901PIR5902 COR5
9
PIR6001PIR6002 COR6
0
PITP1401 COTP
14PITP1501
COTP15
PITP1601 COTP
16PITP1701
COTP17
PIU1901
PIU1902
PIU1903
PIU1904
PIU1905
PIU1906
PIU1907
PIU1908
PIU1909
PIU19010
PIU19011
PIU19012
PIU19013
PIU19014
PIU19015
PIU19016
PIU190EPCO
U19
PIU2001
PIU2002
PIU2003
PIU2004
PIU2005CO
U20
PIU2101
PIU2102
PIU2103
PIU2104
PIU2105
PIU2106
PIU2107
PIU2108
PIU2109
PIU21010
PIU21011
PIU21012
PIU21013
PIU21014
PIU21015
PIU21016
PIU210EPCO
U21
PIU2201
PIU2202
PIU2203
PIU2204
PIU2205CO
U22
PIJP
802
PIJP
902
PIR4801 PIR5302PITP1501
PIU19012
PIU2001
NLDC30EN
PODC30EN
PIJP803
PIU2005
NLDC3M0VOUT
PIR5501 PIR6002PITP1701
PIU21012
PIU2201
NLDC40EN
PODC40EN
PIJP903
PIU2205
NLDC4M0VOUT
PIC5902PIC6002
PIC6202PIC6302
PIC6402PIC6502
PIC6602
PIC6701
PIC6802PIC6902
PIC7202PIC7302
PIC7502PIC7602
PIC7702PIC7802
PIC7902
PIC8001
PIC8102PIC8202
PIR5102PIR5201
PIR5301
PIR5802PIR5901
PIR6001
PIU1906
PIU19010
PIU19015
PIU19016
PIU190EP
PIU2003
PIU2004
PIU2106
PIU21010
PIU21015
PIU21016
PIU210EP
PIU2203
PIU2204
NLGND
PIC5
701
PIC5
802
PIL401
PIU1901
PIU1902
PIC5
801
PIU1903
PIC6102PIR4702 PIR5101
PITP1401PIU1909
PIC6702
PIU1904
PIC6801PIU19011
PIC6901
PIR4
902
PIU1905
PIC7
001
PIC7
102
PIL501
PIU2101
PIU2102
PIC7
101
PIU2103
PIC7402PIR5402 PIR5801
PITP1601PIU2109
PIC8002
PIU2104
PIC8101PIU21011
PIC8201
PIR5
602
PIU2105
PIR5
002
PIU1908
PIR5202
PIU1907
PIR5
702
PIU2108
PIR5902
PIU2107
PIC5901PIC6001
PIC6601
PIC7201PIC7301
PIC7901
PIR4802 PIR5502
PIU19013
PIU19014
PIU2002
PIU21013
PIU21014
PIU2202
NLVI
N0BC
KP
PIC5
702
PIC6101PIC6201
PIC6301PIC6401
PIC6501
PIJP801
PIL402
PIR4701
PIR4
901
PIR5
001
NLVO
UT05
V0BA
CKUP
0LOC
AL
PIC7
002
PIC7401PIC7501
PIC7601PIC7701
PIC7801
PIJP901
PIL502
PIR5401
PIR5
601
PIR5
701
NLVOUT05V0BACKUP0SERVO
PODC30EN
PODC40EN
146
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
Com
pany
:A
Gen
eral
Not
es
Gen
eral
not
es re
latin
g to
the
desi
gn.
Not
e N
o.D
escr
iptio
n
PCB
Lay
out R
equi
rem
ents
Layo
ut n
otes
refe
rring
to sp
ecifi
c co
mpo
nent
s or
gro
up o
f
[Num
ber]
Des
crip
tion
com
pone
nts
are
indi
cate
d on
the
sche
mat
ics
by th
e no
te
PCB
Ass
embl
y V
aria
tions
List
of a
ssem
bly
varia
tions
. Var
iant
s ar
e de
sign
ated
with
a le
tter
Var
iant
Des
crip
tion
star
ting
from
A. T
he d
escr
iptio
n sh
ows
the
Var
iant
nam
e in
num
ber,
encl
osed
in b
race
s, a
djac
ent t
o th
e co
mpo
nent
. Eg
[1].
Not
es
brac
kets
and
then
any
com
men
ts.
A[w
/o h
eade
rs] A
ll SM
T co
mpo
nent
s so
lder
ed.
[1]
GPS
ant
enna
trac
es re
quire
50o
hm
Cop
yrig
ht s
ectio
n (if
app
licab
le)
Not
es.S
chD
oc
2.0
20-D
EC-2
015
1416
Eng.
Jor
ge M
arin
(mar
inm
ar@
ualb
erta
.ca)
UA
AR
G A
utop
ilot.P
rjPc
b
impe
danc
e.
Thro
ugh-
hole
hea
ders
are
not
incl
uded
.
B[w
/ hea
ders
] All
SMT
com
pone
nts
sold
ered
,in
clud
ing
thro
ugh-
hole
hea
ders
.
C[A
ssem
bled
] All
com
pone
nts
are
sold
ered
and
exte
rnal
com
pone
nts
are
plug
ged.
Fully
ass
embl
ed b
oard
.
147
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
Com
pany
:A
SPI2NA
ME
MPU
-600
0
PER
IPH
ER
AL
PB12
MC
U C
S PI
N
MS5
611-
01BA
03PC
13
SPI S
LA
VE
S A
ND
"C
HIP
SE
LE
CT
" PI
N M
APP
ING
Doc
umen
tatio
n.Sc
hDoc
2.0
20-D
EC-2
015
1516
Eng.
Jor
ge M
arin
(mar
inm
ar@
ualb
erta
.ca)
UA
AR
G A
utop
ilot.P
rjPc
b
UAR
T1
NA
ME
TX
FUN
CTI
ON
PA9
MC
U P
IN
RXPA
10
UA
RT
PO
RT
S M
APP
ING
-DE
FAU
LT
SE
T
RC R
ecei
ver 1
TX
UAR
T2TX
PA2
RXPA
3
XBee
RX
XBee
TX
UAR
T3TX
PC10
RXPC
11
GPS
RX
GPS
TX
UAR
T5TX
PC12
RXPD
2
GPI
O P
C12
RC R
ecei
ver 2
TX
UAR
T6TX
PC6
RXPC
7
SERV
O_1
SERV
O_2
UAR
T4TX
PA0
RXPA
1
SERV
O_5
SERV
O_6
148
11
22
33
44
DD
CC
BB
AA
Shee
t Titl
e:
Proj
ectT
itle:
Size
:N
umbe
r:R
ev:
Dat
e:Sh
eet
of
Dra
wn
by:
Com
pany
:A
09-D
EC-2
015
DA
TE
1.0
RE
VIS
ION
RE
VIS
ION
HIS
TO
RY
Initi
al R
elea
se
DE
SCR
IPT
ION
Rev
isio
n H
isto
ry.S
chD
oc
2.0
20-D
EC-2
015
1616
Eng.
Jor
ge M
arin
(mar
inm
ar@
ualb
erta
.ca)
UA
AR
G A
utop
ilot.P
rjPc
b
149
4.5 Autopilot Board Bill-of-Materials (BOM)
This section contains the list of components used in the design of board, including manufacturer and partnumbers.
150
Designator Quantity Value Manufacturer ManufacturerPartNumber
C1,C2 2 10pF Yageo CC0402JRNPO9BN100
C3,C4,C5,C23,C36,C49,C67,C80 8 2.2uF Yageo CC0402MRX5R5BB225
C6,C24 2 4.7uF Yageo CC0603KRX5R6BB475C7,C8,C9,C10,C12,C13,C15,C16,C19,C20,C22,C35,C48,C52,C53,C54,C55,C56,C66,C79,C83 21 100nF Yageo CC0402KRX7R6BB104
C11 1 1uF Yageo CC0402KRX5R5BB105
C14 1 2.2nF Yageo CC0402KRX7R9BB222
C17,C38,C51,C69,C82 5 4.7uF Yageo CC0402MRX5R6BB475
C18 1 220nF Yageo CC0402KRX5R5BB224
C21 1 33uF Kemet T491A336K010AT
C25,C37,C50,C68,C81 5 10nF Yageo CC0402KRX7R7BB103
C27,C40,C58,C71 4 0.47uF Yageo CC0402ZRY5V7BB474
C28,C41,C59,C72 4 10uF SamsungElectro-MechanicsAmerica,Inc. CL32B106KBJNNNE
C29,C42,C60,C73 4 4.7uF SamsungElectro-MechanicsAmerica,Inc. CL32B475KBUYNNE
C30,C43,C61,C74 4 100pF Yageo CC0402JRNPO9BN101
C31,C32,C44,C45,C62,C63,C75,C76 8 47uF MurataElectronicsNorthAmerica GRM32ER71A476KE15L
C33,C46,C64,C77 4 10uF Yageo CC0805KKX5R7BB106
C34,C47,C65,C78 4 100nF Yageo CC0603KPX7R7BB104
C85,C86 2 8.2pF Yageo CC0402BRNPO9BN8R2
C87 1 11mF SeikoInstruments CPH3225A
C88 1 33nF SamsungElectro-MechanicsAmerica,Inc. CL05B333JO5NNNC
C89 1 470pF TDKCorporation C1005C0G1H471J050BA
D1,D2 2 ToshibaSemiconductorandStorage CMS04(TE12L,Q,M)
D3 1 DiodesIncorporated 1N5819HW-7-F
D4 1 MicroCommercialCo 1N4148WX-TP
EX1,EX2 2 SullinsConnectorSolutions NPPN101BFCN-RC
J6,J11,J12,J13,J14,J20 6 Molex,LLC 0533980471
J16 1 3M 961106-5500-AR-PR
J17 1 HiroseElectricCoLtd U.FL-R-SMT(10)
J18 1 CinchConnectivitySolutionsJohnson 142-0701-801
J19 1 Molex,LLC 0533980771
J21 1 Molex,LLC 0533980571
JTAG1 1 SamtecInc. FTSH-105-01-F-DV-K
L1 1 MurataElectronicsNorthAmerica BLM15AX102SN1D
L2,L3,L4,L5 4 6.8uH VishayDale IHLP3232DZER6R8M11
L6 1 6.8nH MurataElectronicsNorthAmerica LQW15AN6N8H00D
L7 1 27nH MurataElectronicsNorthAmerica LQW15AN27NJ00D
LED1,LED5 2 Lite-OnInc. LTST-C190KRKT
LED2 1 Lite-OnInc. LTST-C190AKT
LED3,LED6,LED8,LED9 4 Lite-OnInc. LTST-C190KGKT
LED4,LED7 2 Lite-OnInc. LTST-C190YKT
Q1 1 DiodesIncorporated MMDT5551-7-F
Q2 1 NXPSemiconductors BSS138PS,115R1,R5,R9,R17,R26,R33,R40,R41,R46,R49,R56,R66,R67 13 0 Yageo RC0402JR-070RL
R2 1 9.1K Yageo RC0603FR-079K1L
R3 1 2.2K Yageo RC0603FR-072K2L
R6,R10 2 1K Yageo RC0402FR-071KL
R7,R8,R11,R12,R15,R23,R42,R43,R44,R45,R65 11 10K Yageo RC0402FR-0710KL
R13,R14 2 2.2K Yageo RC0402FR-072K2L
R16 1 120 Yageo RC0402FR-07120RL
R18 1 470 Yageo RC0402JR-07470RL
151
R22,R27,R34,R50,R57 5 100K Yageo RC0402FR-07100KL
R24,R31,R47,R54 4 1Meg RohmSemiconductor MCR01MRTF1004
R25,R32,R48,R55 4 453K Yageo RC0603FR-07453KL
R28,R35,R51,R58 4 249K RohmSemiconductor MCR01MRTF2493
R30,R37,R53,R60 4 100K Yageo RC0603FR-07100KL
R38,R39 2 200K Yageo RC0402FR-07200KL
R61 1 100 Yageo RC0402FR-07100RL
R68 1 10 Yageo RC0402FR-0710RL
RN1,RN2,RN8,RN9 4 1K Yageo YC164-JR-071KL
RN3 1 10K Yageo YC164-JR-0710KL
RN4,RN5 2 2.2K Yageo YC164-JR-072K2L
RN6,RN7 2 470 Yageo YC164-JR-07470RL
S1 1 C&KComponents KMR221GLFS
U1 1 STMicroelectronics STM32F405RGT6
U2 1 MeasurementSpecialties,Inc. MS5611-01BA03
U3 1 InvenSense MPU-6000
U4 1 HoneywellMicroelectronics&PrecisionSensors HMC5883L-TR
U5 1 MaximIntegrated MAX3051EKA+T
U6 1 TexasInstruments LM3940IMPX-3.3/NOPB
U7 1 TexasInstruments LP2992AIM5-3.3/NOPB
U8,U10,U19,U21 4 TexasInstruments LM43603PWPT
U13,U14 2 TexasInstruments PCA9306DCTR
U15,U16 2 TexasInstruments SN74LVC125APWR
U17 1 GLOBALCONNECTORTECHNOLOGY MEM2051-00-195-00-A
U18 1 Atmel ATMEGA328P-AU
U23 1 u-blox MAX-7Q
U24 1 MurataElectronicsNorthAmerica SF1186K-3
U25 1 MaximIntegrated MAX2659ELT+T
USB1 1 Molex,LLC 0473460001
Y1 1 AbraconLLC ABM8G-12.000MHZ-4Y-T3
Y2 1 MurataElectronicsNorthAmerica CSTCE16M0V53-R0
Z1,Z2 2 BournsInc. CG0603MLC-05LE
152
5 Revision HistoryTable 5.1: Falcon I Autopilot Board User Guide Revision History.
Rev. Date Changes1.0 04-Aug-2016 First issue.1.1 29-Nov-2016 Changed board name, updated introduction and updated jumper
table to indicate default jumpers. Added note on solder jumperssection.
1.2 04-Jan-2017 Fixed typographic errors and updated IMU and GPS paragraphs.1.3 25-Jan-2017 Updated cover page picture and added board pictures section.1.4 15-Mar-2018 Added text link to STM32 MCU datasheet. Updated block diagram.1.5 29-Mar-2018 Updated authors. Added BOM.
153
154
Appendix B
Camera Bridge Board Schematics
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
Ind
ex.
Sch
Doc
1.0
14-
DEC
-201
71
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
U_
Doc
ume
ntatio
nD
ocu
ment
atio
n.Sc
hDoc
U_
Revi
sion
Hist
ory
Revi
sion
Hist
ory.
Sch
Doc
Cam
era
US
B H
UB.
Prj
Pcb
1.0
Proj
ect
Titl
e:
Revi
sion
:Da
te:
14-
DE
C-20
17
Bloc
ksP
age
Name
......
.....
...
......
.....
........
........
.....
......
......
......
......
......
......
......
......
...
1Inde
x
2Ca
mera
US
B H
UB
3 4 5 6 7 8 9 18
..............................................................
17
..............................................................
16
..............................................................
15
..............................................................
14
..............................................................
23
..............................................................
22
..............................................................
21
..............................................................
20
..............................................................
19
..............................................................
25
..............................................................
24
..............................................................
11
Not
es10 13
Revi
sion
His
tory
12
Doc
ume
ntation
27
..............................................................
26
..............................................................
28
..............................................................
U_
Came
ra
USB
H
UB
Came
ra
USB
H
UB.
Sch
Doc
..............................................................
U_
Not
esNot
es.Sc
hDoc
US
B H
UB
Came
ra
Arr
ay
US
B S
wit
ch
Came
ra
Conne
ction
Po
wer
Trigge
r Cont
rolle
r
FT
DI
1 5 5
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
Ca
mera
US
B H
UB.
Sch
Doc 1.
014-
DEC
-201
72
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
MS
TT
RG1
MS
TT
RG2
MS
TT
RG3
MS
TT
RG4
MS
TT
RG5
MS
TT
RG6
TX
DR
XD
DT
R
Trigg
er C
ont
roll
erTrigg
er C
ont
roll
er.Sc
hDoc
US
B2US
B3US
B4
US
B1US
BU
US
B H
UB
US
B H
UB.
Sch
Doc
US
B1
US
B2
US
B3
US
B4
US
B5
US
B6
US
B7
US
B8
US
B9
US
B10
US
B11
US
B12
TRI
GG
ER1
TRI
GG
ER2
TRI
GG
ER3
TRI
GG
ER4
TRI
GG
ER5
TRI
GG
ER6
TRI
GG
ER7
TRI
GG
ER8
TRI
GG
ER9
TRI
GG
ER1
0
TRI
GG
ER1
1
TRI
GG
ER1
2
Came
ra
Arra
yCa
mera
Ar
ray.
Sch
Doc
US
B2US
B3US
B4
US
B1US
BU
US
B H
UB
US
B H
UB.
Sch
Doc US
B2US
B3US
B4US
B1US
BU
US
B H
UB
US
B H
UB.
Sch
Doc
MS
TT
RG1
MS
TT
RG2
MS
TT
RG3
MS
TT
RG4
MS
TT
RG5
MS
TT
RG6
US
B
US
B1
US
B2
US
B S
witc
h A
US
B S
witc
h.Sc
hDoc
US
B
US
B1
US
B2
US
B S
witc
h B
US
B S
witc
h.Sc
hDoc
US
B
US
B1
US
B2
US
B S
witc
h C
US
B S
witc
h.Sc
hDoc
US
BU1
US
BU2
US
BU3
US
B2US
B3US
B4
US
B1US
BU
US
B H
UB
US
B H
UB.
Sch
Doc
USBA1USBA2USBA3
D+
D-
US
B
GN
D
US
BV
CC1
D+
D-
US
B
GN
D
US
BV
CC2
D+
D-
US
B
GN
D
US
BV
CC3
D+
D-
US
B
GN
D
US
BV
CC4
US
BC1
US
BC2
US
BC3
US
BC4
US
BC5
US
BC6
US
BC7
US
BC8
US
BC9
US
BC1
0
US
BC1
1
US
BC1
2
CMS
04(
TE1
2L,
Q,M)
D1
US
BV
CC1
MF
-NS
MF2
00-2
CT-
ND
F1
CMS
04(
TE1
2L,
Q,M)
D2
US
BV
CC2
MF
-NS
MF2
00-2
CT-
ND
F2
CMS
04(
TE1
2L,
Q,M)
D3
US
BV
CC3
MF
-NS
MF2
00-2
CT-
ND
F3
CMS
04(
TE1
2L,
Q,M)
D4
US
BV
CC4
MF
-NS
MF2
00-2
CT-
ND
F4
+5
V
Po
wer
Po
wer.
Sch
Doc
VB
US
1
D-2
D+
3
ID
4
GN
D5
SH
SH
P1
Mini
USB
B_1 V
BUS
1
D-2
D+
3
ID
4
GN
D5
SH
SH
P2
Mini
USB
B_1 V
BUS
1
D-2
D+
3
ID
4
GN
D5
SH
SH
P3
Mini
USB
B_1 V
BUS
1
D-2
D+
3
ID
4
GN
D5
SH
SH
P4
Mini
USB
B_1
DT
R
RX
TX
CTS
FT
DIF
TDI.
Sch
Doc
GN
D
PI
D1
01
P
ID
10
2
COD1
PI
D2
01
P
ID
20
2
COD2
PI
D3
01
P
ID
30
2
COD3
PI
D4
01
P
ID
40
2
COD4
PI
F1
01
P
IF
10
2
COF1
PI
F2
01
P
IF
20
2
COF2
PI
F3
01
P
IF
30
2
COF3
PI
F4
01
P
IF
40
2
COF4
PI
P1
01
PI
P1
02
PI
P1
03
PI
P1
04
PI
P1
05
PI
P1
0S
H
CO
P1
PI
P1
01
PI
P1
02
PI
P1
03
PI
P1
04
PI
P1
05
PI
P1
0S
H
PI
P2
01
PI
P2
02
PI
P2
03
PI
P2
04
PI
P2
05
PI
P2
0S
H
CO
P2
PI
P2
01
PI
P2
02
PI
P2
03
PI
P2
04
PI
P2
05
PI
P2
0S
H
PI
P3
01
PI
P3
02
PI
P3
03
PI
P3
04
PI
P3
05
PI
P3
0S
H
CO
P3
PI
P3
01
PI
P3
02
PI
P3
03
PI
P3
04
PI
P3
05
PI
P3
0S
H
PI
P4
01
PI
P4
02
PI
P4
03
PI
P4
04
PI
P4
05
PI
P4
0S
H
CO
P4
PI
P4
01
PI
P4
02
PI
P4
03
PI
P4
04
PI
P4
05
PI
P4
0S
H
PI
D1
02
PI
D2
02
PI
D3
02
PI
D4
02
PI
P1
04
PI
P1
05
PI
P1
0S
H
PI
P2
04
PI
P2
05
PI
P2
0S
H
PI
P3
04
PI
P3
05
PI
P3
0S
H
PI
P4
04
PI
P4
05
PI
P4
0S
H
NLMS
TTRG
1 NL
MSTT
RG2
NLMS
TTRG
3 NL
MSTT
RG4
NLMS
TTRG
5 NL
MSTT
RG6
PI
D1
01
P
IF
10
2
PI
D2
01
P
IF
20
2
PI
D3
01
P
IF
30
2
PI
D4
01
P
IF
40
2
PI
P1
02
PI
P1
03
PI
P2
02
PI
P2
03
PI
P3
02
PI
P3
03
PI
P4
02
PI
P4
03
NLUSBA1 NLUSBA1
NLUSBA2 NLUSBA2
NLUSBA3 NLUSBA3
NLUS
BC1
NLUS
BC1 NL
USBC
2 NL
USBC
2
NLUS
BC3
NLUS
BC3
NLUS
BC4
NLUS
BC4
NLUS
BC5
NLUS
BC5
NLUS
BC6
NLUS
BC6
NLUS
BC7
NLUS
BC7
NLUS
BC8
NLUS
BC8
NLUS
BC9
NLUS
BC9
NLUS
BC10
NL
USBC
10
NLUS
BC11
NL
USBC
11
NLUS
BC12
NL
USBC
12
NLUS
BU1
NLUS
BU1
NLUS
BU2
NLUS
BU2
NLUS
BU3
NLUS
BU3
PI
F1
01
PI
P1
01
PI
F2
01
PI
P2
01
PI
F3
01
PI
P3
01
PI
F4
01
P
IP
40
1
1 5 6
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
US
B H
UB.
Sch
Doc
1.0
14-
DEC
-201
73
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
VSS
1
XO
UT
2
XI
N3
DM4
4
DP4
5
DM3
6
DP3
7
DM2
8
DP2
9
DM1
10
DP1
11
VD
18_
O12
VD33
13
RE
XT
14D
MU
15DP
U16
XRS
TJ17
VB
US
M18
BUSJ
19V
DD5
20V
D33_
O21
DR
V22
LE
D1/
EES
CL
23LE
D2
24P
WRJ
25O
VCJ
26TES
TJ/EES
DA
27V
D18
28U1
FE1.
1S
12
Y1
12
MHz
GN
DUS
B4_
NUS
B4_P
US
B3_
NUS
B3_P
US
B2_
NUS
B2_P
US
B1_
NUS
B1_P
10uF
C8
GN
D
10uF
C9
GN
D
2.7
K1
%
R6
GN
D
US
BU
_N
US
BU
_P
100
K
R4100
K
R2
10uF
C6
10uF
C5+5
V
GN
DG
ND
10uF
C1
GN
D
100
KR5
GN
DG
ND
+5
V
DR
VLE
D1
LE
D2
SD
A3
WP
5
VSS
2
SC
L1
VC
C4
U2
24LC
02B
T-I
/O
T
GN
D
TES
TJ
TES
TJLE
D1
+5
V
GN
D
LE
D5
Red
1K
R7
LE
D2
Gre
en
100nF
C710
K
R3
10nF
C410
K
R1
LE
D4
Gre
en1
KR8
GN
D
LE
D1
Gre
en LE
D3
Gree
n1
KR9
D+D-
US
B
US
B1
D+D-
US
B
US
B2
D+D-
US
B
US
B3
D+D-
US
B
US
B4
D+
D-
US
B
US
BU
LE
D1
LE
D2
DR
V
EE
PR
OM (o
ptio
nal)
VD18
VD18
VD33
VD33
VD33
VD33
VD33
20pF
C2
GN
DG
ND
20pF
C3PIC101 PIC102
CO
C1
PIC201 PIC202
CO
C2
PIC301 PIC302
CO
C3
PIC401 PIC402
CO
C4
PIC5
01
PIC5
02 C
OC
5
PIC6
01
PIC6
02 C
OC
6
PIC701 PIC702 C
OC
7
PIC801 PIC802 C
OC
8
PIC901 PIC902 C
OC
9
PILE
D101
PI
LED1
02
CO
LE
D1
PILE
D201
PI
LED2
02
CO
LE
D2
PILE
D301
PI
LED3
02
CO
LE
D3
PILE
D401
PI
LED4
02
CO
LE
D4
PILED501 PILED502 C
OL
ED
5
PIR1
01
PIR1
02 C
OR
1
PIR2
01
PIR2
02
CO
R2
PIR3
01
PIR3
02 C
OR
3
PIR4
01
PIR4
02
CO
R4
PIR501 PIR502 CO
R5
PIR601 PIR602 CO
R6
PIR701 PIR702 CO
R7
PIR801 PIR802 CO
R8
PIR901 PIR902 CO
R9
PI
U1
01
PI
U1
02
PI
U1
03
PI
U1
04
PI
U1
05
PI
U1
06
PI
U1
07
PI
U1
08
PI
U1
09
PI
U1
01
0
PI
U1
01
1
PI
U1
01
2
PI
U1
01
3
PI
U1
01
4
PI
U1
01
5
PI
U1
01
6
PI
U1
01
7
PI
U1
01
8
PI
U1
01
9
PI
U1
02
0
PI
U1
02
1
PI
U1
02
2
PI
U1
02
3
PI
U1
02
4
PI
U1
02
5
PI
U1
02
6
PI
U1
02
7
PI
U1
02
8
CO
U1
PI
U2
01
PI
U2
02
PI
U2
03
PI
U2
04
PI
U2
05
CO
U2
PI
Y1
01
P
IY
10
2
CO
Y1
PIC6
02
PIR1
01
PIR3
01
PI
U1
02
0
PILE
D101
PI
LED2
01
PILE
D302
PI
LED4
02
PILED501
PI
U1
02
2
NLDR
V
PIC101 PIC201
PIC302
PIC402
PIC5
01
PIC6
01 PIC702
PIC801 PIC901
PIR501
PIR601
PIR701
PI
U1
01
PI
U2
02
PIR901
PI
U1
02
3
PI
U2
01
NLLE
D1
PIR801
PI
U1
02
4
NLLE
D2
PIC202
PI
U1
03
PI
Y1
01
PIC301
PI
U1
02
PI
Y1
02
PIC401 PI
R102
PI
U1
02
6
PIC701 PI
R302
PIR502
PI
U1
01
8
PILE
D102
PILE
D301
PIR902
PILE
D202
PILE
D401
PIR802
PILED502 PIR702
PIR2
01
PI
U1
01
9
PIR4
01
PI
U1
01
7
PIR602
PI
U1
01
4
PI
U1
02
5
PI
U1
02
7
PI
U2
03
NLTE
STJ
PI
U1
01
0
NLUS
B10N
POUSB1
PI
U1
01
1
NLUS
B10P
POUSB1
PI
U1
08
NLUS
B20N
P
OU
SB
2
PI
U1
09
NLUS
B20P
P
OU
SB
2
PI
U1
06
NLUS
B30N
P
OU
SB
3
PI
U1
07
NLUS
B30P
P
OU
SB
3
PI
U1
04
NLUS
B40N
P
OU
SB
4
PI
U1
05
NLUS
B40P
P
OU
SB
4
PI
U1
01
5
NLUS
BU0N
P
OU
SB
U
PI
U1
01
6
NLUS
BU0P
P
OU
SB
U
PIC102
PIC802
PI
U1
01
2
PI
U1
02
8
NLVD
18
PIC5
02
PIC902
PIR2
02
PIR4
02
PI
U1
01
3
PI
U1
02
1
PI
U2
04
PI
U2
05
NLVD
33
POUSB1
POUS
B10D
0
PO
US
B2
PO
USB2
0D0
PO
US
B3
PO
USB3
0D0
PO
US
B4
PO
USB4
0D0
PO
US
BU
PO
USBU
0D0
1 5 7
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
US
B S
witc
h.S
ch
Doc
1.0
14-
DEC
-201
74
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
GN
D
+3
V3
D+
D-
US
B
US
BD+D-
US
B
US
B1
D+D-
US
B
US
B2
GN
D
1 2
P5
Hea
der
2
+3
V3 10
KR1
0
GN
D
100nF
C10
GN
D
1D+
11
D-2
2D+
32
D-4
GN
D5
OE
6
D-7
D+
8
S9
VC
C10
EP
11
U3
TS
3USB2
21E
DR
CR
US
B_P
US
B_
NUS
B1_P
US
B1_
N
US
B2_P
US
B2_
N
PIC1001 PIC1002 C
OC
10
PI
P5
01
PI
P5
02
CO
P5
PIR1001 PIR1002 CO
R1
0
PI
U3
01
PI
U3
02
PI
U3
03
PI
U3
04
PI
U3
05
PI
U3
06
PI
U3
07
PI
U3
08
PI
U3
09
PIU3010
PIU3011
CO
U3
PIC1001 PIR1001
PIU3010
PIC1002
PI
P5
02
PI
U3
05
PI
U3
06
PIU3011
PI
P5
01
PIR1002
PI
U3
09
PI
U3
02
NLUS
B10N
P
OU
SB
1
PI
U3
01
NLUS
B10P
P
OU
SB
1
PI
U3
04
NL
USB2
0N
PO
US
B2
P
IU
30
3
NLUS
B20P
P
OU
SB
2
PI
U3
07
NL
USB0
N P
OU
SB
P
IU
30
8
NLUS
B0P
PO
US
B
PO
US
B
PO
US
B1
PO
USB1
0D0
PO
US
B2
PO
USB2
0D0
POUS
B0D0
1 5 8
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
Ca
mera
Arra
y.S
ch
Doc
1.0
14-
DEC
-201
75
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
Repe
at(
USB
_N)
Repe
at(
USB
_P)
Repe
at(
TRI
GG
ER)
Repe
at(
Came
ra,1
,12
)Ca
mera
Con
necti
on.
Sch
Doc
USP
_N[
1..12
]US
B_
N
USP
_P[1..1
2]
USB_P
US
B_
N1
US
B_
N2
US
B_
N3
US
B_
N4
US
B_
N5
US
B_
N6
US
B_
N7
US
B_
N8
US
B_
N9
US
B_
N10
US
B_
N11
US
B_
N12
US
B_P1
US
B_P2
US
B_P3
US
B_P4
US
B_P5
US
B_P6
US
B_P7
US
B_P8
US
B_P9
US
B_P
10
US
B_P
11
US
B_P
12
D+ D-
US
BUS
B_P1
US
B_
N1
US
B1
D+ D-
US
BUS
B_P2
US
B_
N2
US
B2
D+ D-
US
BUS
B_P3
US
B_
N3
US
B3
D+ D-
US
BUS
B_P4
US
B_
N4
US
B4
D+ D-
US
BUS
B_P5
US
B_
N5
US
B5
D+ D-
US
BUS
B_P6
US
B_
N6
US
B6
D+ D-
US
BUS
B_P7
US
B_
N7
US
B7
D+ D-
US
BUS
B_P8
US
B_
N8
US
B8
D+ D-
US
BUS
B_P9
US
B_
N9
US
B9
D+ D-
US
BUS
B_P
10US
B_
N10
US
B10
D+ D-
US
BUS
B_P
11US
B_
N11
US
B11
D+ D-
US
BUS
B_P
12US
B_
N12
US
B12
TRI
GG
ER
TRIGGER[1..12]
TRI
GG
ER1
TRI
GG
ER2
TRI
GG
ER3
TRI
GG
ER4
TRI
GG
ER5
TRI
GG
ER6
TRI
GG
ER7
TRI
GG
ER8
TRI
GG
ER9
TRI
GG
ER1
0TRI
GG
ER1
1TRI
GG
ER1
2
TRI
GG
ER1
TRI
GG
ER2
TRI
GG
ER3
TRI
GG
ER4
TRI
GG
ER5
TRI
GG
ER6
TRI
GG
ER7
TRI
GG
ER8
TRI
GG
ER9
TRI
GG
ER1
0TRI
GG
ER1
1TRI
GG
ER1
2
NLUS
B0N
NLUS
B0N1
PO
US
B1
NLUS
B0N
NLUS
B0N2
PO
US
B2
NLUS
B0N
NLUS
B0N3
PO
US
B3
NLUS
B0N
NLUS
B0N4
PO
US
B4
NLUS
B0N
NLUS
B0N5
PO
US
B5
NLUS
B0N
NLUS
B0N6
POUSB6
NLUS
B0N
NLUS
B0N7
PO
US
B7
NLUS
B0N
NLUS
B0N8
PO
US
B8
NLUS
B0N
NLUS
B0N9
PO
US
B9
NLUS
B0N
NLUS
B0N1
0
POUSB10
NLUS
B0N
NLUS
B0N1
1 POUSB11
NLUS
B0N
NLUS
B0N1
2
POUSB12
NLUS
B0P
NLUS
B0P1
PO
US
B1
NLUS
B0P
NLUS
B0P2
P
OU
SB
2
NLUS
B0P
NLUS
B0P3
PO
US
B3
NLUS
B0P
NLUS
B0P4
PO
US
B4
NLUS
B0P
NLUS
B0P5
PO
US
B5
NLUS
B0P
NLUS
B0P6
POUSB6
NLUS
B0P
NLUS
B0P7
PO
US
B7
NLUS
B0P
NLUS
B0P8
PO
US
B8
NLUS
B0P
NLUS
B0P9
PO
US
B9
NLUS
B0P
NLUS
B0P1
0
POUSB10
NLUS
B0P
NLUS
B0P1
1
POUSB11
NLUS
B0P
NLUS
B0P1
2
POUSB12
NLTRIGGER0100120 NLTR
IGGE
R1
NLTR
IGGE
R
PO
TR
IG
GE
R1
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R2
PO
TR
IG
GE
R2
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R3
PO
TR
IG
GE
R3
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R4
PO
TR
IG
GE
R4
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R5
PO
TR
IG
GE
R5
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R6
PO
TR
IG
GE
R6
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R7
PO
TR
IG
GE
R7
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R8
PO
TR
IG
GE
R8
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R9
PO
TR
IG
GE
R9
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R10
PO
TR
IG
GE
R1
0
NLTRIGGER0100120 NLTR
IGGE
R
NLTR
IGGE
R11
POTRIGGER11
NLTRIGGER0100120 NLTR
IGGE
R
NLTRIGGER12
POTRIGGER12
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0N0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
NLUS
P0P0
1001
20
PO
TR
IG
GE
R1
P
OT
RI
GG
ER
2
PO
TR
IG
GE
R3
P
OT
RI
GG
ER
4
PO
TR
IG
GE
R5
P
OT
RI
GG
ER
6
PO
TR
IG
GE
R7
P
OT
RI
GG
ER
8
PO
TR
IG
GE
R9
P
OT
RI
GG
ER
10
POTRIGGER11
POTRIGGER12
PO
US
B1
PO
USB1
0D0
PO
US
B2
PO
USB2
0D0
PO
US
B3
PO
USB3
0D0
PO
US
B4
PO
USB4
0D0
PO
US
B5
PO
USB5
0D0
POUSB6
POUS
B60D
0
PO
US
B7
PO
USB7
0D0
PO
US
B8
PO
USB8
0D0
PO
US
B9
PO
USB9
0D0
POUSB10
POUS
B100
D0
POUSB11
POUS
B110
D0
POUSB12
POUS
B120
D0
1 5 9
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
Ca
mera
Conn
ecti
on.
Sch
Doc 1.0
14-
DEC
-201
76
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
125
3
64U4
US
BL
C6-
2S
C6
VB
US
1
D-2
D+
3
ID
4
GN
D5
SH
SH
P11
US
B Mi
ni-
B
+5
V
GN
D
100uF
C11
GN
D
+5
VG
ND
US
B_
N
US
B_P
US
B Conne
ctor
Tri
gger
Co
nne
ctor
TRI
GG
ER
1 2 3 4
P6
Hea
der
4
GN
D
US
BIN
_P
US
BIN
_N
US
B_P
US
B_
NPIC1101 PIC1102 C
OC
11
PI
P6
01
PI
P6
02
PI
P6
03
PI
P6
04
CO
P6
PI
P1
10
1
PI
P1
10
2
PI
P1
10
3
PI
P1
10
4
PI
P1
10
5
PI
P1
10
SH
CO
P1
1
PI
U4
01
PI
U4
02
PI
U4
03
P
IU
40
4
PI
U4
05
PI
U4
06
CO
U4
PIC1101
PI
P1
10
1
PI
U4
05
PIC1102
PI
P6
01
PI
P1
10
4
PI
P1
10
5
PI
P1
10
SH
PI
U4
02
PI
P6
03
PI
P6
04
PI
P6
02
P
OT
RI
GG
ER
PI
P1
10
2
PI
U4
04
NL
USB0
N
PI
P1
10
3
PI
U4
06
NLUS
B0P
PI
U4
03
NL
USBI
N0N
PO
US
B0
N
PI
U4
01
NL
USBI
N0P
PO
US
B0
P
PO
TR
IG
GE
R
PO
US
B0
N
PO
US
B0
P
1 6 0
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
Trigge
r Con
trolle
r.Sc
hDo
c 1.0
14-
DEC
-201
77
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
PC6
(R
ES
ET/
PCI
NT
14)
29
PD0
(R
XD/
PCI
NT1
6)30
PD1
(T
XD/
PCI
NT
17)
31
PD2
(IN
T0/PCI
NT
18)
32
PD4
(PC
IN
T20
/XC
K/T
0)2
VC
C6
GN
D5
PB6
(PC
IN
T6/
XT
AL1
/T
OSC1
)7
PB7
(PC
IN
T7/
XT
AL2
/T
OSC2
)8
PD5
(PC
IN
T21
/OC0
B/T
1)9
PD6
(PC
IN
T22
/OC0
A/A
IN0
)10
PD7
(PC
IN
T23
/AI
N1)
11
PB0
(PC
IN
T0/
CL
KO/
ICP1
)12
PB1
(PC
IN
T1/
OC1
A)13
PB2
(PC
IN
T2/
SS/
OC1
B)14
PB3
(PC
IN
T3/
OC2
A/M
OS
I)15
PB4
(PC
IN
T4/
MIS
O)16
PB5
(S
CK/
PCI
NT
5)17
AV
CC
18
AR
EF
20
GN
D21
PC0
(A
DC0/
PCI
NT
8)23
PC1
(A
DC1/
PCI
NT
9)24
PC2
(A
DC2/
PCI
NT
10)
25
PC3
(A
DC3/
PCI
NT
11)
26
PC4
(A
DC4/
SD
A/P
CIN
T12
)27
PC5
(A
DC5/
SC
L/P
CIN
T13
)28
GN
D3
VC
C4
AD
C619
AD
C722
PD3
(PC
IN
T19
/OC2
B/I
NT1
)1
U5
AT
mega
328P-
AU
GN
D
GN
D
SPI_SC
KSPI_
MIS
OSPI_
MOSI
SPI_
CS
RES
ET
TX
DR
XD
MC
U
DT
R
12
16
MHz
Y2
GN
D
12
34
56
P8
ICSP
RES
ET
SPI_SC
KSPI_
MIS
OSPI_
MOSI
GN
D
+5
V
+5
V
+5
V
+5
V
ICS
P
10
KR1
1
MS
TT
RG1
MS
TT
RG2
MS
TT
RG3
MS
TT
RG4
MS
TT
RG5
MS
TT
RG6
Ju
mpe
rs
12
34
56
78
910
1112
P7
Hea
der 6
X2
MS
TT
RG1
MS
TT
RG2
MS
TT
RG3
MS
TT
RG4
MS
TT
RG5
MS
TT
RG6
1K
R12 LE
D6
Gre
en
GN
D
100nF
C13
100nF
C15
100nF
C14
100nF
C12
MS
TT
RG1
MS
TT
RG2
MS
TT
RG3
MS
TT
RG4
MS
TT
RG5
MS
TT
RG6
TX
DR
XD
DT
RPI
C120
1 PI
C120
2
COC12
PIC1301 PIC1302 COC13
PIC1401 PIC1402 COC14
PIC1501 PIC1502 COC15
PILED601 PILED602 CO
LED6
PI
P7
01
P
IP
70
2
PI
P7
03
P
IP
70
4
PI
P7
05
P
IP
70
6
PI
P7
07
P
IP
70
8
PI
P7
09
PIP7010
PIP7011
PIP7012
COP7
PI
P8
01
P
IP
80
2
PI
P8
03
P
IP
80
4
PI
P8
05
P
IP
80
6
COP8
PIR1101 PIR1102 COR11
PIR1201 PIR1202 COR12
PI
U5
01
PI
U5
02
PI
U5
03
PI
U5
04
PI
U5
05
PI
U5
06
PI
U5
07
PI
U5
08
PI
U5
09
PIU5010
PIU5011
PIU5012
PIU5013
PIU5014
PIU5015
PIU5016
PIU5017
PIU5018
PIU5019
PIU5020
PIU5021
PIU5022
PIU5023
PIU5024
PIU5025
PIU5026
PI
U5
02
7
PIU5028
PI
U5
02
9
PI
U5
03
0
PIU5031
PIU5032
COU5
PIY2
01
PIY202 PIY203
COY2
PIC1301 PIC1401
PIC1501
PI
P8
02
PIR1101
PI
U5
04
PI
U5
06
PIU5018
PIU5020
PIC1
202
NLDT
R P
OD
TR
PIC1302 PIC1402
PIC1502
PILED602
PI
P8
06
PI
U5
03
PI
U5
05
PIU5021
PIY203
PI
P7
02
P
IP
70
1
PIU5032
POMSTTRG1
NLMS
TTRG
1
PI
P7
04
P
IP
70
3
PI
U5
01
PO
MS
TT
RG
2
NLMS
TTRG
2
PI
P7
06
P
IP
70
5
PI
U5
02
PO
MS
TT
RG
3
NLMS
TTRG
3
PI
P7
08
P
IP
70
7
PI
U5
09
PO
MS
TT
RG
4
NLMS
TTRG
4
PIP7010
PI
P7
09
PIU5010
PO
MS
TT
RG
5
NLMS
TTRG
5
PIP7012
PIP7011
PIU5011
PO
MS
TT
RG
6
NLMS
TTRG
6
PILED601 PIR1201
PI
U5
07
PIY2
01
PI
U5
08
PIY202
PIU5012
PIU5013
PIU5019
PIU5022
PIU5023
PIU5024
PIU5025
PIU5026
PI
U5
02
7
PIU5028
PIC1
201
PI
P8
05
PIR1102
PI
U5
02
9
NLRE
SET
PI
U5
03
0
NLRX
D P
OR
XD
PIU5014
NLSP
I0CS
PI
P8
01
PIU5016
NLSP
I0MI
SO
PI
P8
04
PIU5015
NLSP
I0MO
SI
PI
P8
03
PIR1202 PIU5017
NLSP
I0SC
K
PIU5031
NLTX
D P
OT
XD
PO
DT
R
POMSTTRG1
PO
MS
TT
RG
2
PO
MS
TT
RG
3
PO
MS
TT
RG
4
PO
MS
TT
RG
5
PO
MS
TT
RG
6
PO
RX
D
PO
TX
D
1 6 1
11
22
33
44
DD
CC
BB
AA
She
et:
Pro
ject:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
FT
DI.
Sch
Doc
1.0
14-
DEC
-201
78
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
VC
CIO
4
VC
C20
US
BDP
15US
BD
M16
NC
8
RES
ET
19
NC
24
OS
CI27
OS
CO
28
3V3
OU
T17
AGND25
GND7
GND18
GND21
TEST26
TX
D1
RX
D5
RTS
3
CTS
11
DT
R2
DS
R9
DC
D10
RI6
CB
US0
23
CB
US1
22
CB
US2
13
CB
US3
14
CB
US4
12
U8
FT23
2R
L SS
OP
GN
D
GN
D
GN
DG
ND
US
B5
V
+5
V
+5
V
+5
V+5
VDT
R
RX
TX
1.5
KR2
11.
5K
R20
CTS
FT
DI_3
V3
LE
D8
Red
LE
D9
Red
10uF
C30
100nF
C31
100nF
C32
VB
US
1
D-2
D+
3
ID
4
GN
D5
SH
SH
P12
Mini
USB
B_2
CMS
04(
TE1
2L,
Q,M)
D5
US
B5
V+5
V
+5
V
PIC3001 PIC3002 COC30
PIC3101 PIC3102 COC31
PIC3201 PIC3202 COC32
PI
D5
01
P
ID
50
2
COD5
PILED801 PILED802 CO
LED8
PILED901 PILED902 CO
LED9
PI
P1
20
1
PI
P1
20
2
PI
P1
20
3
PI
P1
20
4
PI
P1
20
5
PI
P1
20
SH
COP12
PIR2001 PIR2002 COR20
PIR2101 PIR2102 COR21
PI
U8
01
PI
U8
02
PI
U8
03
PI
U8
04
PI
U8
05
PI
U8
06
PIU807
PI
U8
08
PI
U8
09
PI
U8
01
0
PI
U8
01
1
PI
U8
01
2
PI
U8
01
3
PI
U8
01
4
PI
U8
01
5
PI
U8
01
6
PI
U8
01
7
PIU8018
PI
U8
01
9
PI
U8
02
0
PIU8021
PI
U8
02
2
PI
U8
02
3
PI
U8
02
4
PIU8025 PIU8026
PI
U8
02
7
PI
U8
02
8 COU8
PIC3002 PIC3101
PI
D5
02
PIR2002 PIR2102
PI
U8
04
PI
U8
02
0
PI
U8
01
1
PO
CT
S
PI
U8
02
P
OD
TR
PIC3201 P
IU
80
17
PIC3001 PIC3102
PIC3202
PI
P1
20
5
PI
P1
20
SH
PIU807 PIU8018
PIU8021 PIU8025
PIU8026
PILED801 PIR2001 PILED802 P
IU
80
22
PILED901 PIR2101 PILED902 P
IU
80
23
PI
P1
20
2
PI
U8
01
6
PI
P1
20
3
PI
U8
01
5
PI
P1
20
4
PI
U8
03
PI
U8
06
PI
U8
08
PI
U8
09
PI
U8
01
0
PI
U8
01
2
PI
U8
01
3
PI
U8
01
4
PI
U8
01
9
PI
U8
02
4
PI
U8
02
7
PI
U8
02
8
PI
U8
05
P
OR
X
PI
U8
01
P
OT
X
PI
D5
01
PI
P1
20
1
PO
CT
S
PO
DT
R
PO
RX
P
OT
X
1 6 2
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
Po
wer.
Sch
Doc
1.0
14-
DEC
-201
79
12Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
SW
1
SW
2
CB
OO
T3
VC
C4
BIAS
5
SY
NC
6
RT
7
PG
OO
D8
FB
9
AG
ND
10
SS/
TR
K11
EN
AB
LE
12
VI
N13
VI
N14
PG
ND
15
PG
ND
16
PA
D
U6
LM
43603P
WP
T
GN
DG
ND
10uF
C17
453
KR1
4
100
KR1
8
GN
D
4.7uF
C18
GN
DG
ND
GN
D
VI
N
TP2
0.47uF
C16
10nF
C26
GN
D
IH
LP3
232
DZER
6R8
M11
L1 6.8u
H
4.7uF
C27
GN
D
1Me
g
R13
100pF
C20
249
KR1
7
GN
D
100
K
R16
47uF
C21
10uF
C23
100nF
C24
GN
DG
ND
GN
DG
ND
+5
V
2.2uF
C25
0R15
47uF
C22
2.2
V Th
resho
ld
DC1
_E
N
TP1
100nF
C19
GN
D
GN
DG
ND
GN
D
IN
3
1
OU
T2
GN
D
U7
NCP
1117
LPS
T33
T3G
+3
V3
+5
V+5
V
10uF
C28
10uF
C29
1K
R19
LE
D7
Gre
en
Buc
k Co
nver
ter
LD
O 3.
3V
Regul
ator
1 2
P10
Hea
der
2
1 23
J1
PWR
JA
CK
VI
N
GN
DG
ND
VIN
Po
wer
Jac
k and
Hea
ders
1 2
P9
Hea
der
2
GN
D
+5
V
PIC1
601
PIC1
602
COC16
PIC1701 PIC1702 COC17
PIC1801 PIC1802 COC18
PIC1901 PIC1902 COC19
PIC2001 PIC2002 COC20
PIC2101 PIC2102 COC21
PIC2201 PIC2202 COC22
PIC2301 PIC2302 COC23
PIC2401 PIC2402 COC24
PIC2501 PIC2502 COC25
PIC2601 PIC2602 CO
C26
PIC2701 PIC2702 CO
C27
PIC2801 PIC2802 COC28
PIC2901 PIC2902 COC29
PIJ1
01
PIJ1
02
PIJ1
03
COJ1
PIL101
PIL102
COL1
PILED701 PILED702 CO
LED7
PI
P9
01
PI
P9
02
COP9
PIP1001
PIP1002
COP10
PIR1301 PIR1302 COR13
PIR1401 PIR1402 COR14
PIR1
501
PIR1
502 COR15
PIR1
601
PIR1
602 COR16
PIR1701 PIR1702 COR17
PIR1801 PIR1802 COR18
PIR1901 PIR1902 COR19
PITP101 COTP1
PITP
201
COTP2
PI
U6
01
PI
U6
02
PI
U6
03
PI
U6
04
PI
U6
05
PI
U6
06
PI
U6
07
PI
U6
08
PI
U6
09
PI
U6
01
0
PI
U6
01
1
PI
U6
01
2
PI
U6
01
3
PI
U6
01
4
PI
U6
01
5
PI
U6
01
6
PI
U6
0E
P COU6
PIU701
PI
U7
02
P
IU
70
3
PI
U7
04
COU7
PIC2901
PI
U7
02
P
IU
70
4
PIC2001 PIC2101
PIC2201 PIC2301
PIC2401
PIC2801 PIL102
PI
P9
01
PIR1301
PIR1
501
PIR1
601
PIR1902 P
IU
70
3
PIR1401 PIR1802
PITP
201
PI
U6
01
2
NLDC
10EN
PIC1702
PIC1802 PIC1902
PIC2102 PIC2202
PIC2302 PIC2402
PIC2501
PIC2602 PIC2702
PIC2802 PIC2902
PIJ1
02
PIJ1
03
PILED702
PI
P9
02
PIP1002
PIR1702
PIR1801
PI
U6
06
PI
U6
01
0
PI
U6
01
5
PI
U6
01
6
PI
U6
0E
P
PIU701
PIC1
601
PI
U6
03
PI
C160
2 PIL
101
PI
U6
01
PI
U6
02
PIC2002
PIR1302 PIR1701
PITP101 P
IU
60
9
PIC2502
PI
U6
04
PIC2601 P
IU
60
11
PIC2701
PIR1
502
PI
U6
05
PILED701 PIR1901
PIR1
602
PI
U6
08
PI
U6
07
PIC1701 PIC1801
PIC1901
PIJ1
01
PIP1001
PIR1402
PI
U6
01
3
PI
U6
01
4
1 6 3
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
MH1
MH2
MH3
MH4
Mou
ntin
g Hol
es 3.
175
mm
(125
mil) p
ad,
5.714
mm
(225
mil) d
rill
BO
AR
D M
OU
NTI
NG
HO
LES
Mounting
Hol
es
Fiduc
ials
FI
D1
FID2
FID
UCI
ALS
3X
TOP
FI
D3
Gene
ral
Note
s
Gene
ral no
tes
rela
ting to
the
de
sign.
Not
e No.
Desc
riptio
n
PC
B La
yout
Requi
reme
nts
Layo
ut no
tes
refe
rrin
g to
spe
cifi
c co
mpo
nent
s or gr
oup
of
[Nu
mbe
r]De
scriptio
n
compo
nent
s ar
e in
dica
ted
on th
e sc
hema
tics
by th
e no
te
PC
B A
sse
mbl
y Va
riations
List
of a
sse
mbl
y va
riatio
ns.
Vari
ants
are
des
igna
ted
with
a l
ette
r
Vari
ant
Desc
riptio
n
star
ting
fro
m A.
The
desc
riptio
n sho
ws t
he
Vari
ant
name
in
nu
mbe
r, e
nclo
sed in
br
aces
, ad
jace
nt t
o th
e co
mpo
nent.
Eg [1].
Mecha
nica
l
brac
kets
and th
en a
ny
com
ment
s.
A[Pr
otot
ype
A]
Stra
ight
mini
USB
con
nect
ors.
[1]
90oh
m i
mpe
danc
e for
USB
di
ffer
enti
al pa
irs.
Copy
righ
t se
ctio
n (if ap
plic
able
)
Not
es.Sch
Doc
1.0
14-
DEC
-201
710
12
Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
[Pr
otot
ype
B]
R/A
mini
USB
con
nect
ors.
B
COFI
D1
COFI
D2
COFI
D3
PIMH101
COMH
1
PIMH201
COMH
2
PIMH301
COMH
3
PIMH401
COMH
4
PIMH101 PIMH201
PIMH301 PIMH401
1 6 4
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
Doc
ume
ntation.
Sch
Doc
1.0
14-
DEC
-201
711
12
Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
1 6 5
11
22
33
44
DD
CC
BB
AA
She
et
Titl
e:
Pro
ject
Titl
e:
Size
:N
umbe
r:Re
v:Da
te:
She
et of
Dra
wn by
:
Co
mpa
ny:
A
14-
DE
C-101
7
DA
TE
1.0
RE
VISI
ON
RE
VISI
ON
HIS
TO
RY
Firs
t re
leas
e.
DES
CR
IP
TIO
N
Revi
sion
Hist
ory.
Sch
Doc
1.0
14-
DEC
-201
712
12
Eng. J
orge
Ma
rin
(jl
mari
nma
rcan
o@g
mail.c
om)
Ca
mera
US
B H
UB.
Prj
Pcb
1 6 6
167
Appendix C
Camera Bridge BoardBill-of-Materials
Designator Quantity Value Manufacturer ManufacturerPartNumberC1_USBHUB1,C1_USBHUB2,C1_USBHUB3,C1_USBHUB4,C5_USBHUB1,C5_USBHUB2,C5_USBHUB3,C5_USBHUB4,C6_USBHUB1,C6_USBHUB2,C6_USBHUB3,C6_USBHUB4,C8_USBHUB1,C8_USBHUB2,C8_USBHUB3,C8_USBHUB4,C9_USBHUB1,C9_USBHUB2,C9_USBHUB3,C9_USBHUB4,C30 21 10uF Yageo CC0603MRX5R5BB106C2_USBHUB1,C2_USBHUB2,C2_USBHUB3,C2_USBHUB4,C3_USBHUB1,C3_USBHUB2,C3_USBHUB3,C3_USBHUB4 8 20pF Yageo CC0603JRNPO9BN200C4_USBHUB1,C4_USBHUB2,C4_USBHUB3,C4_USBHUB4 4 10nF Yageo CC0603KRX7R8BB103C7_USBHUB1,C7_USBHUB2,C7_USBHUB3,C7_USBHUB4,C10_USBSwitchA,C10_USBSwitchB,C10_USBSwitchC,C12,C13,C14,C15,C24,C31,C32 14 100nF Yageo CC0603KRX7R9BB104C11_Camera1,C11_Camera2,C11_Camera3,C11_Camera4,C11_Camera5,C11_Camera6,C11_Camera7,C11_Camera8,C11_Camera9,C11_Camera10,C11_Camera11,C11_Camera12 12 100uF MurataElectronics GRM31CR60J107ME39KC16 1 0.47uF Yageo CC0402ZRY5V7BB474
C17 1 10uFSamsungElectro-MechanicsAmerica,Inc. CL32B106KBJNNNE
C18 1 4.7uFSamsungElectro-MechanicsAmerica,Inc. CL32B475KBUYNNE
C19 1 100nF Yageo CC0402KRX7R6BB104C20 1 100pF Yageo CC0402JRNPO9BN101
C21,C22 2 47uFMurataElectronicsNorthAmerica GRM32ER71A476KE15L
C23,C28,C29 3 10uF Yageo CC0805ZKY5V6BB106C25 1 2.2uF Yageo CC0402MRX5R5BB225C26 1 10nF Yageo CC0402KRX7R7BB103C27 1 4.7uF Yageo CC0402MRX5R6BB475
D1,D2,D3,D4,D5 5ToshibaSemiconductorandStorage CMS04(TE12L,Q,M)
F1,F2,F3,F4 4 BournsInc. MF-NSMF200-2J1 1 CUI-Inc. PJ-075CH-SMT-TRL1 1 6.8uH VishayDale IHLP3232DZER6R8M11
168
LED1_USBHUB1,LED1_USBHUB2,LED1_USBHUB3,LED1_USBHUB4,LED2_USBHUB1,LED2_USBHUB2,LED2_USBHUB3,LED2_USBHUB4,LED3_USBHUB1,LED3_USBHUB2,LED3_USBHUB3,LED3_USBHUB4,LED4_USBHUB1,LED4_USBHUB2,LED4_USBHUB3,LED4_USBHUB4,LED6,LED7 18 Kingbright APT1608SGCLED5_USBHUB1,LED5_USBHUB2,LED5_USBHUB3,LED5_USBHUB4,LED8,LED9 6 Kingbright APT1608SURCKP1,P2,P3,P4,P12 5 TEConnectivityAMP 1734510-1
P6_Camera1,P6_Camera2,P6_Camera3,P6_Camera4,P6_Camera5,P6_Camera6,P6_Camera7,P6_Camera8,P6_Camera9,P6_Camera10,P6_Camera11,P6_Camera12 12 Molex,-LLC 530480410P11_Camera1,P11_Camera2,P11_Camera3,P11_Camera4,P11_Camera5,P11_Camera6,P11_Camera7,P11_Camera8,P11_Camera9,P11_Camera10,P11_Camera11,P11_Camera12 12 HiroseElectric UX20-MB-5PR1_USBHUB1,R1_USBHUB2,R1_USBHUB3,R1_USBHUB4,R3_USBHUB1,R3_USBHUB2,R3_USBHUB3,R3_USBHUB4,R10_USBSwitchA,R10_USBSwitchB,R10_USBSwitchC,R11 12 10K Yageo AC0603FR-0710KLR2_USBHUB1,R2_USBHUB2,R2_USBHUB3,R2_USBHUB4,R4_USBHUB1,R4_USBHUB2,R4_USBHUB3,R4_USBHUB4,R5_USBHUB1,R5_USBHUB2,R5_USBHUB3,R5_USBHUB4,R18 13 100K Yageo RC0603FR-07100KLR6_USBHUB1,R6_USBHUB2,R6_USBHUB3,R6_USBHUB4 4 2.7K Yageo RC0603FR-072K7LR7_USBHUB1,R7_USBHUB2,R7_USBHUB3,R7_USBHUB4,R8_USBHUB1,R8_USBHUB2,R8_USBHUB3,R8_USBHUB4,R9_USBHUB1,R9_USBHUB2,R9_USBHUB3,R9_USBHUB4,R12,R19 14 1K Yageo RC0603FR-071KLR13 1 1Meg RohmSemiconductor MCR01MRTF1004R14 1 453K Yageo RC0603FR-07453KLR15 1 0 Yageo RC0402JR-070RLR16 1 100K Yageo RC0402FR-07100KLR17 1 249K RohmSemiconductor MCR01MRTF2493R20,R21 2 1.5K Yageo RC0603FR-071K5L
169
U1_USBHUB1,U1_USBHUB2,U1_USBHUB3,U1_USBHUB4 4 TerminusTech FE1.1sU3_USBSwitchA,U3_USBSwitchB,U3_USBSwitchC 3 TexasInstruments TS3USB221EDRCR
U4_Camera1,U4_Camera2,U4_Camera3,U4_Camera4,U4_Camera5,U4_Camera6,U4_Camera7,U4_Camera8,U4_Camera9,U4_Camera10,U4_Camera11,U4_Camera12 12 STMicroelectronics USBLC6-2SC6U5 1 Atmel ATMEGA328P-AUU6 1 TexasInstruments LM43603PWPTU7 1 ONSemiconductor NCP1117LPST33T3G
U8 1FTDI,FutureTechnologyDevicesInternationalLtd FT232RL-REEL
Y1_USBHUB1,Y1_USBHUB2,Y1_USBHUB3,Y1_USBHUB4 4 ECSInternational ECS-120-20-33-TR
Y2 1MurataElectronicsNorthAmerica CSTCE16M0V53-R0
170
171
Appendix D
3D Printer Settings
The following table summarizes the settings used for 3D printing the PLA parts.
Table D.1: 3D printer settings.
Setting Units Value
Layer Height mm 0.18First Layer Height mm 0.27Perimeter Shells 2Top Solid Layers 3Bottom Solid Layers 3Fill Density % 15-20Fill Pattern Hexagonal (Honeycomb)Combine Infill Every 2 LayersPrint Speed mm/s 60Travel Speed mm/s 80Extruder Temperature °C 200Platform Temperature °C 50