Image Processing Reference
historical data through change detection techniques (Berger 1998 ). Mobile GIS is
also a growing technology for field spatial data collection such as LBS (Location Based
Services), which provides GIS and spatial data information via mobile and field units.
The growing use of GIS in crime mapping has also increased public access to
crime data, especially through the most accessible and popular medium, the World
Wide Web. Moreover, complex GIS applications can lead to scenario planning to
develop hazard plans and public safety for floodwater management, earthquake or
other disasters. GIS can also be relied upon as an effective rescue effort to assist fire
and police departments. It will also be particularly useful for code enforcement
officials who wish to locate property owners.
Dubai Police Practice 1
With the launching of e-government initiative in Dubai/UAE, technologies such as
remote sensing and GIS can clearly contribute to all phases of the city's initiative
to improve the command level decision-making in the Dubai Government. This
section presents some of the efforts being taken in this regard by Dubai Police.
Image Processing of IKONOS Data of Dubai
Figure 16.4 shows a snapshot of one of the IKONOS images acquired for a portion
of Dubai City covering the area from Dubai's airport to Dubai Creek. Panchromatic
and multi-spectral IKNONS data were fused together based on the Hue Saturation
Value (HSV). In this image sharpening technique, the higher resolution panchro-
matic data sharpens the detail of the multi-spectral data by transforming the RGB
colors to an HSV color space. During this transformation process, band values in
the low-resolution multi-spectral are replaced by values from the high-resolution
panchromatic images based on the hue and saturation properties of the multi-
spectral data. The result is a multi-spectral image with a higher resolution than that
of the original raw data.
Figure 16.5 shows the different steps involved in the HSV transformation pro-
cess. To get a better image with the pixel size of a high-resolution image but with-
out signs of image “blockness,” a bilinear re-sampling (BL) technique was applied
to the HSV-sharpened image. The resolution of this image is one meter; however,
at certain conditions of lighting and background color what is depicted by such
images may surpass the theoretical limits of the ground resolution. For example, in
Fig. 16.5c , striations less than one meter in width allow for the identification of the
direction in which the grass of the golf course had been mowed.
1 The figures shown in this chapter are for illustrative purposes only; they do not constitute legal
or official documentation.