A hybrid manmade-target detection algorithm for polarimetric sar images

This paper proposes a hybrid target detection algorithm for polarimetric synthetic aperture radar (SAR). The proposed algorithm includes three parts: Span Detector (SD), Polarimetric Whitening Filter Detector (PWF) and Optimal Polarimetric Detector (OPD). The detection output of SD is used to optimize the input of PWF, and the detection output of PWF is used to optimize the input of OPD. The iteration is also employed in the proposed method. A new criterion is presented to compare the performances of different polarimetric target detection algorithms. According to the new criterion, different detection algorithms will select the same number of pixels as their detection results. Experimental results demonstrate that the performance of the proposed algorithm is better than other algorithms, and the new criterion is more appropriate for the comparison of different polarimetric target-detection algorithms.