Spatial Classification Techniques for Synthetic Aperture Radar

In this paper, we examine new methods for the classification of single frequency C-band Synthetic Aperture Radar (SAR) data using trigonometric functions to segment the information into primary band composites. The decomposition is shown to be scaleable from the complex series representation to a simple addition of real and imaginary terms. A condensed example is provided for four filter elements in the standard orientation. This example shows the spatial and radiometric classification of a meandering river system using the radiometric and spatial filters. A discussion of the primary effects is given with a complete derivation for the filter design, the transfer functions, and the related scalar applications.