QUANTITATIVE COMPARISON OF THE PERFORMANCE OF SAR SEGMENTATION ALGORITHMS

Methods to evaluate the performance of segmentation algorithms for syn thetic radar (SAR) images are developed, based on known properties of coherent speckle and a scene model in which areas of constant backscat ter coefficient are separated by abrupt edges. Local and global measur es of segmentation homogeneity are derived and applied to the outputs of two segmentation algorithms developed for SAR data, one based on it erative edge detection and segment growing, the other based on global maximum a posteriori (MAP) estimation using simulated annealing. The q uantitative statistically based measures appear consistent with visual impressions of the relative quality of the segmentations produced by the two algorithms. On simulated data meeting algorithm assumptions, b oth algorithms performed well but MAP methods appeared visually and me asurably better. On real data, MAP estimation was markedly the better method and retained performance comparable to that on simulated data, while the performance of the other algorithm deteriorated sharply. Imp rovements in the performance measures will require a more realistic sc ene model and techniques to recognize oversegmentation.