Joint Polarimetric Subspace Detector Based on Modified Linear Discriminant Analysis

Polarimetric synthetic aperture radar (PolSAR) is widely used in remote sensing and has important applications the detection of ships. Although many polarimetric detectors have been proposed, they are not well combined. Recently, a optimization filter (PDOF) was which performs most environments. In this study, novel subspace form PDOF [strict (SPDOF)] further developed based on Cauchy inequality matrix decomposition theories, enhancing performance. Furthermore, simple method to determine optimal dimension detector trace ratio proposed by calculating area under receiver operating characteristic (ROC) curve, reaching best performance among subspaces detector. Moreover, combine different detectors, modified linear discriminant analysis for diagonal loading (DLD) subspaces. The experimental results demonstrate superiority these joint detectors. Most importantly, DLD solves previous limitations due complex clutter background achieves comparable that Wishart (Gaussian) distribution, particularly low target-to-clutter (TCR) case.