A Novel Polarimetric CFAR Target Detection Method

A new polarimetric synthetic aperture radar (PolSAR) image CFAR target detector is proposed in this paper. By introducing the inverse Gamma distribution which is extensively used in modeling, the distribution of polarimetric matched filter (PMF) metric, denoted as P_G⁰, is derived on the product model; Furthermore, a fast and exact parameter estimation method of P_G⁰ distribution is presented using the "second kind statistics" based on the Mellin transform; Finally the formula of the CFAR detection threshold is deduced, and the target detection using the proposed constant alarm rate (CFAR) detector is performed on the RADARSAT-2 PolSAR data. Experimental results demonstrate the great efficiency of the P_G distribution and the corresponding parameter estimation method in data fitting of areas with different degree of homogeneity. Moreover, the successive CFAR detector can successfully complete the automatic target detection with low false alarm rate and high detection rate in complex clutter environment where the homogeneity of terrain varies sharply.