detection context of this study, the periphery of the image is
not as important, and brightness fall-off should be consistent
between bi-temporal pairs. Automatic mitigation of camera
vignetting did not produce large or noticeable image artifacts
in the process, and had an effectiveness proportional to the
initial degree of vignetting. When capturing airborne imagery,
selection of an aperture, or aperture range, should be carefully
considered relative to the sensor gain and
shutter speed, and
the collective effect of all three on image acuity and intra-
frame brightness consistency for the ease of
RSI
frame to frame
registration and difference image creation.
Temporal consistency in image brightness was evaluated
relative to how much background image noise was present in
images with different capture parameters. The trends in Figure
7 look very similar to the intensity standard deviation results
in Figures 4a and 4b, except that lower values were preferable
in the former, and higher values preferable in the latter two.
EV
bias values furthest from +0.7
EV
are the best choice for low
noise change detection images, excluding the images with an
exposure bias of >+0.7, which had low image noise stemming
from high numbers of truncated (255 value) pixels. Images
with negative exposure biases greater than −2 did not exhibit
truncated values, but had a small standard deviation of pixel
values and low contrast, which lowered the measured differ-
ence in image noise, as shown in Figures 4a and 4b.
Detection results from images with simulated cracks
exhibit the same optimal
EV
bias as several of the previous
tests, with the values nearest to −1.3 and −1EV, yielding the
best detection results. Small linear features such as cracks in
an asphalt road within the scene (which were present in the
before and after images) were also readily detectable with the
−1.3 and −1EV images.
The experiments designed and executed in this thesis
study were conducted in such a way, that the methods should
be useful for many remote sensing professionals looking to
use consumer
DSLR
cameras. However different scene char-
acteristics, analysis objectives, and sensors, may limit the
applicability of the results discussed here.
Additional work utilizing a similar testing framework, but
a different camera system, and scene, would be helpful in de-
termining which of the results obtained in this study pertain
to specific parameters of the study, and which are more wide-
ly applicable to general end users. Specifically, with regard to
the simulated damage crack detection, the testing of datasets
that include actual damage, and other types of damage such
as rubble and object deformation, may prove insightful. The
capture of images on overcast days, at high or lower latitudes,
and from airborne platforms could also be used to address ad-
ditional questions about the effects of those factors on image
capture best practices.
This study is novel in its focus on the radiometric char-
acteristics of
DSLR
cameras, its ground-based image capture
testing approach, and its analytical damage change detection
context,
DSLR
cameras offer a convenient and low-cost solu-
tion to many remote sensing applications for which a large
format imaging radiometers are beyond users’ operational,
budgetary, and/or technical capabilities. We find that expo-
sure biases slightly less than the auto-exposure selected
EV
bias ±0, are preferable for change detection. Additionally, an
ISO
less than 200 is preferable, with an
ISO
of 100 or less being
ideal. As for the light metering, shutter speed, and aperture,
we did not find an optimal combination of camera settings
which performs best in all scenarios. Instead, we found differ-
ent settings are preferable, depending on the estimated
AIM
of
the collection, the illumination of the scene, and the spectral
characteristics of the objects of interest in the scene.
DSLR
cameras exhibit high radiometric fidelity and can effectively
support low-cost aerial image-based change detection, such as
for post-hazard damage assessment.
References
Ahrends, H.E., R. Brügger, R. Stöckli, J. Schenk, P. Michna, F.
Jeanneret, H. Wanner, and W. Eugster, 2008. Quantitative
phenological observations of a mixed beech forest in northern
Switzerland with digital photography,
Journal of Geophysical
Research
, 113:1–11.
Allen, C.W., 1973.
Astrophysical Quantities
, England, 125 p.
Bockaert, V., 2006. Sensor Sizes, URL:
web/20110415004138/http://www.dpreview.com/learn/?/Glossar
y/Camera_System/sensor_sizes_01.htm
(last date accessed: 05
January 2018).
Clemens, S.R., 2015.
Procedures for Correcting Digital Camera
Imagery Acquired by the AGGIEAIR Remote Sensing Platform
,
Utah State University.
Coulter, L., D. Stow, and S. Baer, 2003. A frame center matching
technique for precise registration of multitemporal airborne
frame imagery,
IEEE Transactions on Geoscience and Remote
Sensing
, 41(11):2436–2444.
Coulter, L., and D. Stow, 2008. Assessment of the spatial co-
registration of multitemporal imagery from large format digital
cameras in the context of detailed change detection,
Sensors
,
8(4):2161–2173.
Coulter, L., D. Stow, Y.H. Tsai, C. Chavis, C. Lippitt, G. Fraley, and
R. McCreight, March 2012. Automated detection of people and
vehicles in natural environments using high temporal resolution
airborne remote sensing,
Proceedings of the ASPRS Annual
Conference
, March 2012, pp. 1–13.
Dean, C., T.A. Warner, and J.B. McGraw, 2000. Suitability of the
DCS460c colour digital camera for quantitative remote sensing
analysis of vegetation,
ISPRS Journal of Photogrammetry and
Remote Sensing
, 55(2):105–118.
DXOMARK, 2012. Sigma 105mm F2.8 EX DG Macro Nikon mounted
on Nikon D800E: Tests and Reviews, URL:
dxomark.com/Lenses/Sigma/Sigma-105mm-F28-EX-DG-Macro-
Nikon-mounted-on-Nikon-D800E__814
(last date accessed: 05
January, 2018).
Green, D.W.B., 1992. Correcting for atmospheric extinction,
International Comet Quarterly
, 14:55–59.
Lebourgeois, V., A. Begue, S. Labbe, B. Mallavan, L. Prevot, and
B. Roux, 2008. Can commercial digital cameras be used as
multispectral sensors?, A crop monitoring test,
Sensors
,
8(11):7300–7322.
Levin, N., E. Ben-Dor, and A. Singer, 2005. A digital camera as a tool
to measure colour indices and related properties of sandy soils
in semi-arid environments,
International Journal of Remote
Sensing
, 26(24):5475–5492.
Richardson, A.D., B.H. Braswell, D.Y. Hollinger, J.P. Jenkins, and
S.V. Ollinger, 2009. Near-surface remote sensing of spatial
and temporal variation in canopy phenology,
Ecological
Applications
, 19(6):1417–1428.
Stow, D., A. Hope, A. Nguyen, S. Phinn, and C. Benkelman, 1996.
Monitoring detailed land surface changes from an airborne
multispectral digital camera system,
IEEE Transactions on
Geoscience and Remote Sensing
, 34:1191–1202
.
Stow, D., L. Coulter, and S. Baer, 2003. A frame centre matching
approach to registration for change detection with fine spatial
resolution multi-temporal imagery,
International Journal of
Remote Sensing
, 24(19):3873–3879.
Stow, D.A., L.L. Coulter, G.R. MacDonald, and C.D. Lippitt, 2016.
Evaluation of geometric capture and processing elements in
the context of a repeat station imaging approach to registration
and change detection,
Photogrammetric Engineering &Remote
Sensing
, 82:775–788.
158
March 2018
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
