12
MULTISCALE FRACTAL
CHARACTERISTICS OF URBAN
LANDSCAPE IN INDIANAPOLIS, USA
BINGQING LIANG AND QIHAO WENG
12.1 INTRODUCTION
A major problem in urban remote sensing is the heterogeneity of the urban environment. For example, low-density residential areas may be composed of tree crowns,
rooftops, lawns, paved streets, driveways, and parking lots. Such a heterogeneous
landscape pattern is usually displayed as a spatial variation of spectral responses and
structural features like texture in remotely sensed images, presenting a challenge for
image analysis and interpretation. In this situation, it is necessary to focus on the
overall spatial pattern of variation that characterizes each urban category. However,
remote sensing analysis based on spatial information in general has not been well
understood. One of the challenges is how to characterize and derive the overall spatial
complexity of different land surface features either individually or as a whole from
images. Ideally, the extracted spatial pattern should represent those that may present at
their operational scale, or the scale at which the features operate in the real world. The
fast-growing remote sensing community has been able to provide numerous images
with varied spatial, spectral, radiometric, and temporal resolutions. These resolutions
can be interpreted as the scale of observation or measurement and they may or may not
match up with the operational scale for a given spatial feature. Therefore, depending
on the scene’s resolutions, an individual object’s spatial pattern may appear heterogeneous at one level but homogeneous at another and is not always comparable to that
in nature (Turner and Garnder, 1991). For a given spatial object, a clear knowledge
230
Scale Issues in Remote Sensing, First Edition. Edited by Qihao Weng.
2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.
MULTISCALE FRACTAL
CHARACTERISTICS OF URBAN
LANDSCAPE IN INDIANAPOLIS, USA
BINGQING LIANG AND QIHAO WENG
12.1 INTRODUCTION
A major problem in urban remote sensing is the heterogeneity of the urban environment. For example, low-density residential areas may be composed of tree crowns,
rooftops, lawns, paved streets, driveways, and parking lots. Such a heterogeneous
landscape pattern is usually displayed as a spatial variation of spectral responses and
structural features like texture in remotely sensed images, presenting a challenge for
image analysis and interpretation. In this situation, it is necessary to focus on the
overall spatial pattern of variation that characterizes each urban category. However,
remote sensing analysis based on spatial information in general has not been well
understood. One of the challenges is how to characterize and derive the overall spatial
complexity of different land surface features either individually or as a whole from
images. Ideally, the extracted spatial pattern should represent those that may present at
their operational scale, or the scale at which the features operate in the real world. The
fast-growing remote sensing community has been able to provide numerous images
with varied spatial, spectral, radiometric, and temporal resolutions. These resolutions
can be interpreted as the scale of observation or measurement and they may or may not
match up with the operational scale for a given spatial feature. Therefore, depending
on the scene’s resolutions, an individual object’s spatial pattern may appear heterogeneous at one level but homogeneous at another and is not always comparable to that
in nature (Turner and Garnder, 1991). For a given spatial object, a clear knowledge
230
Scale Issues in Remote Sensing, First Edition. Edited by Qihao Weng.
2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.