robot sensing and perception - engineeringpetriu/robot-percep_emil.pdf · school of information...
TRANSCRIPT
Emil M. Petriu, Dr. Eng., FIEEEProfessor School of Information Technology and EngineeringUniversity of OttawaOttawa, ON., K1N 6N5 [email protected]
Robot Sensing and Perception
some of the research done in the SMRLab
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Why natural binary coding cannot be used in practice for absolute position recovery ?A n-bit code would be needed for each quantization step, resulting in n binary tracks in parallel with the guide-path. For instance, the encoding of a 160 m long guide-path with a 0.01 m resolution would need 14 tracks running in parallelwith the guide path
0 0 0 0 1 0 1 0 1 1 1 0 1 1 0 0 0 1 1 1 1 1 0 0 1 1 0 1 0 0 1
0 5 10 15 2520 30p =
PRBS=
A (2n -1) term Pseudo-Random Binary Sequences (PRBS) generated by a n-bit modulo-2 feedback shift register is used as an one-bit / quantization-step absolute code. The absolute position identification is based on the PRBS window property. According to this any n-tuple seen through a n-bit window sliding over PRBS is unique and henceforth it fully identifies each position of the window.
The figure shows, as an example, a 31-bit term PRBS: 0, 0, 0, 0, 1, 0, 1, 0, 1, 1, 1, 0, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 1, 1, 0,1, 0, 0, 1, generated by a 5-bit shift register. The 5-bit n-tuples seen through a window sliding over this PRBs are unique and represent a 1-bit wide absolute position code.
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
è pseudo-random encoding provides a practicalsolution allowing absolute position recovery withany desired n-bit resolution while employing onlyone binary track, regardless of the value of n.
m•t-1 m •t .... m •t+r ... m • t+t-1 (m+1) • t (m+1) •t+1
Q(m) Q(m+1)
Positions on the scale
Milestones
{x(k)=S(p+n-k)|k=n,...,1} Pseudo-random n-typle corresponding to the position index p = m•t + r
m•t
Control Logic
Shift
Reverse PRBS feedback logic
R(n) R(1)
Stop count
Counter
SEQUENTIAL CODE CONVERSION
. . .
rp = m•t + r
Natural code corresponding to the current position
Adder
. ..
MS
. ..
...
Q(0)
Q(1)
Q(w)
B(n) B(1)
R(n) R(1)
PARALLEL MILESTONE IDENTIFICATION&CONVERSION
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Serial-parallel pseudo-random / natural code conversion algorithm
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Pseudo-random encoded track (one bit per quantization step) allows recovery of the absolute position of an optically guided Automated Guided Vehicle (AGV)
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Optically guided AGV tracking the pseudo-random encoded track
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Pseudo-random encoded guide path allows recoveryof the absolute position of an AGV using computer vision
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Computer vision recognition of the pseudo-random binary code
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Model-based recognition of a pseudo-random encoded object
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Wall-mounted pseudo-random encoded guide path allows recovery of the absolute position of the AGV using computer vision
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Pseudo-random encoding for computer vision recovery of the 3D position of a probe mapping the electromagnetic–field radiated by a telephone set
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Computer vision recovery of the pseudo-random code
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
0
10
0
0
0
0
0 0
0
1
0
1
0
1
1
1
1
0
0
1
1 1
1 1 1
1
1
1
0 1 1 0 0 1
0 1
1 0
0
0
1
1 0 0 1
0 0
0
1 1
0 1 1 1
0 0
1 10
0 01
j
0 1 2 3 4 5 6 7 8
0
1
2
3
4
5
6
i
Illustrating the window property in a Pseudo-Random Binary Array (PRBA). The 3-by-2 code seen trough a window on a 7-by-9 PRBA is unique and used as absolute code for the window position (i,j).
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Pseudo-Random Binary Array (PRBA) encoding for the recovery of the 2D absolute position of a free ranging mobile robot using computer vision
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Mobile robot navigation using multiple IR sensors and vision
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Path Specifications
ENVIRONMENT
Task
ROBOT
JOINT/WHEEL SERVOS
I.R. RANGE SENSORS
Actuator
Wheel Position
WHEEL/STEER ENCODERS
VISION
ONBOARD COMPUTER
Raster Image
Local Connection
Remote Connection
Object Identities and POSEs
Position Specifications
FRAME TRANSFORMS
TRAJECTORY PLANNER
ROBOT MODELFRAME TRANSFORMS
Robot Position
TASK PLANNER
GEOMETRIC WORLD MODEL
POSITION MODEL
OBJECT RECOGNITION
TRAJECTORY PARAMETRS ESTIMATION
Trajectory Constraints
TELEOPERATOR
VIDEO MONITOR
TAC
TILE
M
ON
ITO
R
&
JOYS
TIC
K
TACTILE SENSOR
COMPOSITE WORLD MODEL
Model-based telepresence control of the mobile robot
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Recovery of the 3D shape of objects using structured light
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
P
P'
P*
PROJECTION GRID
LIGHT SOURCE
CAMERA'S CENTER OF PERSPECTIVE
IMAGE
OBJECT
W
x
y
I
G
x
x
y
y
z
d
CAMERA'S OPTICAL AXIS
PROJECTOR'S OPTICAL AXIS
zz
Q
Point identification problem in pseudo-random encoded structured light
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Pseudo-random color encoded structured-lightgrid projected on a 3D object
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Recovered corner points at the intersection of grid line edges
Pseudo-random color encodedstructured light grid projected on a cube
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Haptic perception of 3D object geometry: the robot arm provides the kinesthetic capability and the tactile sensor probe provides the cutaneous information
Force Sensitive Resistor (FSR) tactile sensor array, 16-by-16 sensing elements on a 1 square inch area
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
Instrumented passive compliant wrist for tactile exploration of objects
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu
“Tactile display” for computer-human interaction allowing humans to feel through their own touch sense computer-generated geometric profiles. It consists of an array of 8-by-8 vibrators on a 1 square inch area.
University of Ottawa School of Information Technology - SITE
Sensing and Modelling Research LaboratorySMRLab - Prof.. Emil M. Petriu