following the described method are plotted in Figure 8. The
concentration of curves in the mid-temperature regions clearly
shows the unseparability of the related classes by using this
feature alone (compare Figure 2).
We did not consider here the surface slopes which affect
the incident solar flux and hence also the surface temper-
ature. Surface slopes lead to misclassification in almost all
albedo based classification applications too (Wiemker 1995),
(Wiemker 1998).
7 CONCLUSIONS
An algorithm has been proposed which separates a given
scene into vegetational and non vegetational regions. It solely
relies on remotely sensed measurements of a part of the di-
urnal temperature curve without using any spectral bands in
the frequency range of reflected sunlight.
As mentioned above the used algorithm cannot compete with
methods based on reflected sunlight such as the NDVI. But
the results clearly recommend the use of thermal data as a
complement to albedo data for classification purposes.
ACKNOWLEDGEMENT
This work was supported by the Volkswagen-Stiftung. The
image flights were conducted in collaboration with the Ger-
man Aerospace Center (DLR WeBling, München), particularly
with the help of Volker Amann, Peter Hausknecht, Rudolf
Richter and Manfred Schrôder.
References
Duda, R. O. and P. E. Hart (1973). Pattern Classification
and Scene Analysis. Wiley, New York, 1973.
Gash, J. H. C. (1987). An analytical framework for extrapo-
lating evaporation measurements by remote sensing sur-
face temperature. International Journal of Remote
Sensing 8 (8), 1245-1249, 1987.
Holbe, H. R. and J. C. Luvall (1989). Modeling Surface Tem-
perature Distributions in Forest Landscapes. Remote
Sensing of Environment 27, 11-24, 1989.
Kim, H. K. (1992). Urban heat island. International Jour-
nal of Remote Sensing 13 (12), 2319-2336, 1992.
Lambin, E. F. and A. H. Strahler (1994). Indicators of land-
cover change for change-vector analysis im multitempo-
ral space at cooarse spatial scales. International Jour-
nal of Remote Sensing 15 (10), 2099-2119, 1994.
Leak, S. M. and G. Venugopal (1990). Thematic Mapper
thermal infrared data in discriminating selected urban
features. International Journal of Remote Sensing
11 (5), 841-857, 1990.
Price, J. C. (1977). Thermal Inertia Mapping: A New View of
the Earth. Journal of Geophysical Research 82 (18),
2582-2590, 1977.
Price, J. C. (1981). The contribution of thermal data in
Landsat multispectral classification. Photogrammetric
Engineering and Remote Sensing 47 (2), 229-236,
1981.
234
Richter, R. (1992). Radiometrische Auslegung von Sensoren
und quantitative Auswertung von Fernerkundungsdaten
im optischen Spektralbereich. Dissertation, German
Aerospace Research Establishment (DLR), Oberpfaffen-
hofen, 1992.
Smith, R. C. G. and B. J. Choudhury (1991). Analysis of nor-
malized difference and surface temperature observations
over southeastern Australia. International Journal of
Remote Sensing 12 (10), 2021-2044, 1991.
Wiemker, R. (1995). Improved Color Constant Classification
of Remotely Sensed Multispectral Imagery. In Stein,
T.l., editor, Proceedings of the International Geoscience
and Remote Sensing Symposium IGARSS 1995, IEEE,
ISBN 0-7803-2567-2, volume 2, pages 1153-1155, 1995.
Wiemker, R. (1996). Registration of Airborne Scanner Im-
agery Using Akima Local Quintic Polynomial Interpo-
lation. In Proceedings of the Second International Air-
borne Remote Sensing Conference and Exhibition, San‘
Francisco, Ann Arbor, 1996, volume Ill, pages 210-219.
Environmental Research Institut of Michigan.
Wiemker, R. (1998). The Color Constancy Problem in Mul-
tispectral Remote Sensing. Dissertation, Universitat
Hamburg, Fachbereich Physik, 1998.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998
