Image Processing Reference
In-Depth Information
Chapter 4
The Spectral Dimension in Urban Remote
Sensing
Martin Herold and Dar A. Roberts
Urban environments are characterized by different types of materials and land
cover surfaces than found in natural landscapes. The analysis of remote sensing
data has to consider these unique spectral characteristics. This chapter describes the
spectral properties of urban areas, how different urban land cover types are spec-
trally discriminated, and which sensor configurations are most useful to map urban
areas. We also show how new remote sensing technologies improve our capabilities
to map urban areas in high spatial and thematic detail.
Learning Objectives
Upon completion of this chapter, you should be able to:
Distinguish the unique spectral characteristics of urban areas
Explain the separability and most suitable spectral bands in
discriminating urban land cover type
Speculate on the potential of hyperspectral, multispectral and
LIDAR remote sensing data in urban mapping
4.1
Introduction
The spectral signal is one of the most important properties of land surfaces
measured with remote sensing (Fig.
4.1
). The amount and spectral qualities of
energy acquired by the remote sensing system (at sensor radiance) is dependent
M. Herold (
*
)
Institute of Geo-Information Science and Remote Sensing, Wageningen University,
Droevendaalsesteeg 3, Gaia, building number 101, P.O. Box 6708, Wageningen, The Netherlands
e-mail: martin.herold@wur.nl
D.A. Roberts
Geography Department, University of California, 5832 Ellison Hall, Santa Barbara,
CA 93106-4060, USA
e-mail: dar@geog.ucsb.edu
