Implementation of a Fast Algorithm for Segmenting SAR Imagery

Abstract

This report gives a detailed presentation of the implementation of a new fast algorithm for image segmentation. The original motivation for development of the algorithm was the segmentation of synthetic aperture radar (SAR) imagery into homogeneous regions for target detection in the Analysts' Detection Support System. However, the algorithm is a general one based upon Mumford-Shah functionals, and there is no technical reason why it could not also be used for other imaging modalities, including multiband imagery. The algorithm has computational complexity on the order of the Fast Fourier Transform, the benchmark for fast algorithms.