do’s for uav spectral remote sensingaasen, h., bolten, a., 2018. multi-temporal high-resolution...
TRANSCRIPT
||Helge Aasen
[email protected]@PhenoFly |
0
Do’s for UAV spectral remote sensing
19.11.2019
Helge Aasen1*, Andreas Bolten2, Georg Bareth2, Eija Honkavaara3, Arko Lucieer4, Pablo Zarco-
Tejada5
1Crop Science Group, Institute of Agricultural Sciences, ETH Zurich, Switzerland2GIS and RS Research Group, Institute of Geography, University of Cologne, Germany3Department of Remote Sensing and Photogrammetry, Finnish Geospatial Research Institute, Finland4Discipline of Geography and Spatial Sciences, School of Technology, Environments and Design, College of Sciences and Engineering, University of Tasmania,
Australia5European Commission (EC), Joint Research Centre (JRC), Italy
||Helge Aasen
[email protected]@PhenoFly |
19.11.2019 1
field
data
pro
duct
Aasen, H., Honkavaara, E., Lucieer, A., Zarco-Tejada, P., 2018. Quantitative Remote Sensing at Ultra-High Resolution
with UAV Spectroscopy: A Review of Sensor Technology, Measurement Procedures, and Data Correction Workflows.
Remote Sensing
particle / object
in environment
‘pixel’ in digital
representation
?
||Helge Aasen
[email protected]@PhenoFly |
19.11.2019 2
field
data
pro
duct
Aasen, H., Honkavaara, E., Lucieer, A., Zarco-Tejada, P., 2018. Quantitative Remote Sensing at Ultra-High Resolution
with UAV Spectroscopy: A Review of Sensor Technology, Measurement Procedures, and Data Correction Workflows.
Remote Sensing
particle / object
in environment
‘pixel’ in digital
representation
?
Outline
1. Radiometric in-field (vicarious) calibration
2. Different data processing schemes give you different results
3. How to keep track of all of that
||Helge Aasen
[email protected]@PhenoFly |
▪ Radiometric calibration determines the transformation
from pixel values to the physically meaningful information
Spectral & radiometric calibration workflow
19.11.2019 3
Laboratory
RC1: relative calibration to uniform response
SC: spectral calibration estimates band configuration
DNn: normalized DN
RC2: absolute radiometric calibration to (at-sensor radiance)
Field
Transformation to reflectance factors
e.g. with the empirical line method (ELM)
Aasen, H., Honkavaara, E., Lucieer, A., Zarco-Tejada, P., 2018. Quantitative Remote Sensing at Ultra-High Resolution with UAV Spectroscopy: A Review of
Sensor Technology, Measurement Procedures, and Data Correction Workflows. Remote Sensing
||Helge Aasen
[email protected]@PhenoFly |
4
Radiometric calibration protocol
Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers – from
theory to application..
||Helge Aasen
[email protected]@PhenoFly |
5
Radiometric calibration protocol
Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers – from
theory to application..
||Helge Aasen
[email protected]@PhenoFly |
Take home message #1
6
▪ Radiometric calibration: Don’t use “field-spectroscopy like
calibration”
➢ Use empirical line + radiometric block adjustment or irradiance
sensor [1, 2]
[1] Aasen, H., Honkavaara, E., Lucieer, A., Zarco-Tejada, P., 2018. Quantitative Remote Sensing at Ultra-High Resolution with UAV Spectroscopy: A Review of
Sensor Technology, Measurement Procedures, and Data Correction Workflows. Remote Sensing
[2] Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers – from theory to application. Remote
Sensing of Environment.
||Helge Aasen
[email protected]@PhenoFly |
I(x,y,λ)
Across-track
Along-track
Ort
ho
recti
ficati
on
via
SfM + GCPs
(and/or imu + gnss)
Drawings kindly provided by
Stefan Livens (VITO)
Aasen, H., Honkavaara, E., Lucieer, A., Zarco-Tejada, P., 2018. Quantitative Remote Sensing at Ultra-High Resolution with UAV
Spectroscopy: A Review of Sensor Technology, Measurement Procedures, and Data Correction Workflows. Remote Sensing
(spectral) 2D imagers2D imagers
||Helge Aasen
[email protected]@PhenoFly |
9Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers –
from theory to application. Remote Sensing of Environment.
Imaging spectroscopy with 2D imagers
19.11.2019
||Helge Aasen
[email protected]@PhenoFly |
11/19/2019 10
Single (most nadir) image
Mosaic, blending: disabled
Mosaic, blending: average
Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers –
from theory to application. Remote Sensing of Environment.
Imaging spectroscopy with 2D imagers
||Helge Aasen
[email protected]@PhenoFly |
11Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers –
from theory to application. Remote Sensing of Environment.
Imaging spectroscopy with 2D imagers
19.11.2019
||Helge Aasen
[email protected]@PhenoFly |
11/19/2019 12
Single (most nadir) image
Mosaic, blending: disabled
Mosaic, blending: average
Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers –
from theory to application. Remote Sensing of Environment.
Imaging spectroscopy with 2D imagers
||Helge Aasen
[email protected]@PhenoFly |
11/19/2019 13Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers –
from theory to application. Remote Sensing of Environment.
A: single image
||Helge Aasen
[email protected]@PhenoFly |
11/19/2019 14Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers –
from theory to application. Remote Sensing of Environment.
A: single imageThe specific field of view is the composition of pixels and
their angular properties within a scene used to
characterize a specific AOI on the ground
➢ Not a UAV problem
➢ Always when we use different sensing settings
||Helge Aasen
[email protected]@PhenoFly |
11/19/2019 15Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging
spectroscopy with hyperspectral 2D imagers – from theory to application. RSE
ASD
Sin
gle
im
ag
eB
len
din
g:
dis
ab
led
Ble
nd
ing
: a
ve
rag
e
Imaging spectroscopy with 2D imagers
||Helge Aasen
[email protected]@PhenoFly |
Different processing modes give different results
▪ Blending disabled:
▪ viewing geometry close to nadir
▪ More proportion of soil
▪ Close to hemispherical directional reflectance factors (HDRF)
▪ Blending mode average
▪ Viewing geometry composed of sensor FOV
▪ Relatively more proportion of plant material
▪ Hemispherical conical reflectance factors (HCRF)
Take home message #2
19.11.2019 16
➢ Be careful to compare apples with apples [2]
[2] Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers
– from theory to application. Remote Sensing of Environment.
||Helge Aasen
[email protected]@PhenoFly |
19.11.2019 17
field
data
pro
duct
particle / object
in environment
‘pixel’ in digital
representation
?
Take home message #3:➢Keep track of your metadata [1,3,4]
[1] Aasen, H., Honkavaara, E., Lucieer, A., Zarco-Tejada, P., 2018. Quantitative Remote Sensing at Ultra-High Resolution with
UAV Spectroscopy: A Review of Sensor Technology, Measurement Procedures, and Data Correction Workflows. Remote
Sensing
[3] Aasen, H., Burkart, A., Bolten, A., Bareth, G., 2015. Generating 3D hyperspectral information with lightweight UAV snapshot
cameras for vegetation monitoring: From camera calibration to quality assurance. ISPRS Journal of Photogrammetry and Remote
Sensing
[4] Roth, L., Hund, A., Aasen, H., 2018. PhenoFly Planning Tool: flight planning for high-resolution optical remote sensing with
unmanned areal systems. Plant Methods
||Helge Aasen
[email protected]@PhenoFly |
▪ Radiometric calibration: Don’t use “field-spectroscopy like
calibration”
➢ Use empirical line or irradiance sensor [1, 2]
➢ Different processing modes
➢ Be careful to compare apples with apples [2]
➢ Traceability of your workflow
➢ Use metadata [1,3,4]
Summary
18
[1] Aasen, H., Honkavaara, E., Lucieer, A., Zarco-Tejada, P., 2018. Quantitative Remote Sensing at Ultra-High Resolution with UAV Spectroscopy: A Review of
Sensor Technology, Measurement Procedures, and Data Correction Workflows. Remote Sensing
[2] Aasen, H., Bolten, A., 2018. Multi-temporal high-resolution imaging spectroscopy with hyperspectral 2D imagers – from theory to application. Remote
Sensing of Environment.
[3] Aasen, H., Burkart, A., Bolten, A., Bareth, G., 2015. Generating 3D hyperspectral information with lightweight UAV snapshot cameras for vegetation
monitoring: From camera calibration to quality assurance. ISPRS Journal of Photogrammetry and Remote Sensing
[4] Roth, L., Hund, A., Aasen, H., 2018. PhenoFly Planning Tool: flight planning for high-resolution optical remote sensing with unmanned areal systems. Plant
Methods
@PhenoFly I www.PhenoFly.net I [email protected]
I www.kp.ethz.ch