A discrimination algorithm and the effect of resolution

Abstract

The discrimination stage of the Lincoln Laboratory multistage target detection algorithm for synthetic aperture radar (SAR) data is examined. This second stage of the multistage algorithm is intended to separate targets and man-made objects from naturally occurring clutter. Results for this algorithm are compared for 1-ft-resolution SAR data that have been processed using the polarimetric whitening filter (PWF) against results using 1-m-resolution SAR data that have been processed in a similar manner. Theoretical expressions are derived. They are shown to accurately predict the performance of the discrimination algorithm. The features used in the discrimination algorithm are described. The feature sets are different for the high- and low-resolution data sets. The finer resolution data are found to provide better performance, i.e., more than an order of magnitude reduction in false alarm rate for the same probability of detection. >