Conclusions
Urban
VIS
signatures extracted from the Hyperion image show
distinct spectral curves at broader level, and show very good
classification accuracy. Inter-class confusion between bare soil
and concrete or residential land use class at times is evident
from results.
VIS
mixture signatures show potential to be used
as a surrogate definition for land use classes such as residen-
tial demographic zones.
Image-based calibration methods such as
FAR
and
IAR
provide many advantages over physics-based models such as
6S/V
. They are simple to implement and all the information
required is available within the image itself. Both these meth-
ods are inherently explorative and sound assessment of merits
and demerits of both are required before wide application. In
case of
IAR
, it is important to know if dominant material in
the scene has adverse effect on the calibration, and in case of
FAR
, it is important to know which flat field might be the best
in a given constraints of the study area. Some of the specific
conclusions and recommendations of our study are:
1. All the reflectance calibration methods capture mixed
signature of pixels with known dominant materials accu-
rately and can be effectively used for impervious surfaces
in absence of a pure material signature. In fact, it reduces
confused results because of bright pure signatures at times.
Furthermore, mixed signatures of
VIS
fractions in a certain
proportions can represent land use classes, for example,
upmarket residential versus low economy residential.
2.
FAR
, on an average, shows marginal improvements in over-
all classification accuracies as compared to
IAR
and other
physics-based methods such as
6S
. The improvement in
User’s accuracy and Producer’s accuracy over other meth-
ods is evident for many of the flat fields. Industrial roof
covers provide one of the best classification results. Inter-
estingly, Plain regions and Industrial Roofs are slightly bet-
ter flat fields than play grounds. Based on the availability
of sufficient number of pixels, any of these three regions
can be effectively used for calibrating a hyperspectral im-
age of an urban area. Further, flat fields reduced confusion
between the bright signatures of soil and concrete.
3.
IAR
and
FAR
produce sufficiently accurate signatures to
detect different types of vegetation covers (for example,
in Experiment 8, related to water hyacinth and tree), but
further experiments with more vegetation types in urban
areas are required to understand the possibilities. The
EO-1
equation shows absorption artifacts and
6S
also shows
some reflection artifacts in the signatures.
6S
signatures ap-
pear to be overcorrected in visible range (Figure 6).
4. In presence of bare soil reference spectra, interclass confu-
sion (with concrete or residential) is reduced as bright
spots in plain region are mistaken to be open ground
instead of concrete roofs. Similarly, some of the bright
spots in the residential area, such as concrete roof tops are
mistaken to be surfaces similar to stone quarry area.
5. Amplitude information is important for impervious surfaces
as spectral shapes of the most of the impervious surfaces are
very similar to each other, and the only difference is in the
amplitude. Techniques like NS
3
(Nidamanuri
et al
., 2011)
that takes into account amplitude information which would
be helpful to discriminate intra impervious classes further.
Acknowledgments
Shailesh Deshpande would like to thank Priya Deshpande for
her timely assistance during the trips conducted for survey-
ing land use land covers for this study. He would like to also
thank Robin Wilson for his support for Py6S interface. We
thank principal investigators for their effort in establishing
and maintaining AERONET Pune site.
References
AERONET - Aerosol robotic network, 2015. URL:
.
nasa.gov/
, NASA Godard Space Center (last date acessesd: 18
March 2017).
Anderson, J.R, E.E. Hardy J.T. Roach, and R.E. Witmer, 1976. Land use
and land cover classification system for use with remote sensor
data,
Geological Survey Professional Paper 964
, United States
Government Printing Office, Washington, D.C.
Arnold, Jr., C.L. and C.J. Gibbons, 1996. Impervious surface coverage:
The emergence of key environmental indicator,
American
Planning Association
, 62(2):243–258.
Barnsley, M.J., and S.L. Barr, 1996. Inferring urban land use from
satellite sensor images using kernel-based spatial reclassification,
Photogrammetric Engineering & Remote Sensing
, 62:949–958.
Beck, R.A., 2003.
EO-1 User Guide v. 2.3
, USGS Earth Resources
Observation Systems Data Center (EDC), Sioux Falls, South
Dakota, 57198–0001.
Benediktsson, J.A., J.A. Palmason, and J.R. Sveinsson, 2005.
Classification of hyperspectral data from urban areas based
on extended morphological profiles,
IEEE Transactions on
Geoscience and Remote Sensing
, 43(3): 480–491.
Biggar, S.F., K.J. Thome, and W. Wisniewski, 2003. Vicarious
radiometric calibration of EO-1 sensors by reference to high-
reflectance ground targets,
IEEE Transactions on Geoscience and
Remote Sensing
, 41(6):1174–1179.
Cadenasso, M.L., S.T.A. Pickett, and K. Schwarz, 2007. Spatial
heterogeneity in urban ecosystems: Reconceptualizing land
cover and a framework for classification,
Frontiers in Ecology
and the Environment
, 5(2):80–88.
Chavez, P.S., Jr., 1988. An improved dark-object subtraction technique
for atmospheric scattering correction of multispectral data,
Remote Sensing of Environment
, 24(3):459– 479.
Chavez, P.S., Jr., 1996. Image-based atmospheric corrections -
Revisited and improved,
Photogrammetric Engineering &
Remote Sensing
, 62(9):1025–1036.
Chen, J., and G.F. Hepner, 2001. Investigation of imaging spectroscopy
for discriminating urban land covers and surface materials,
Proceedings AVIRIS Earth Science and Applications Workshop
,
Palo Alto, California, Jet Propulsion Laboratory, Pasadena,
California,
JPL Publication 02-1
.
Clark, R.N., G.A. Swayze, K.E. Livo, R.F. Kokaly, T.V.V. King, J.B.
Dalton, J.S. Vance, B.W. Rockwell, T. Hoefen, and R.R. McDougal,
2002. Surface reflectance calibration of terrestrial imaging
spectroscopy data: A tutorial using AVIRIS,
Proceedings of the
10
th
Airborne Earth Science Workshop
,
JPL Publication, 02-1
.
Comber, A.J., C.F. Brunsdon, and C.J.Q. Farmer, 2012. Community
detection in spatial networks: Inferring land use from a planar
graph of land cover objects,
International Journal of Applied
Earth Observation and Geoinformation
, 18:274–282.
Conel, J.E., R.O. Green, G. Vane, C.J. Bruegge, and R.E. Alley, 1987.
AIS-2 radiometry and a comparison of methods for the recovery
of ground reflectance,
Proceedings of the 3
rd
Airborne Imaging
Spectrometer Data Analysis Workshop
, (G. Vane, editor)
Pasadena, California, Jet Propulsion Laboratory,
JPL Publication.
87-30:
18–47.
Congalton, R.G., 1991. A review of assessing the accuracy of
classifications of remotely sensed data,
Remote Sensing of
Environment
, 37:35–46.
Deshpande, S.S., A.B. Inamdar, and H.M. Vin, 2014. Improved dark
object technique for atmospheric correction of EO1-Hyperion
data,
Proceedings of GEOMATRIX 2014
, Mumbai, India.
Deshpande, S.S., A.B. Inamdar, and H.M. Vin, 2013. Overview of
hyperspectral remote sensing of impervious surfaces in urban
environment, GRMSE 2013 (F. Bian, Y. Xie, X. Cui, and Y.
Zeng, editors) Springer-Verlag Berlin Heidelberg,
Proceedings of
GRMSE
, Part I, CCIS 398:128-140.
374
May 2017
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
