SCHATTEN MATRIX NORM BASED POLARIMETRIC SAR DATA REGULARIZATION. APPLICATION OVER CHAMONIX MONT-BLANC

The paper addresses the filtering of Polarimetry Syn-thetic Aperture Radar (PolSAR) images. The filtering strat-egy is based on a regularizing cost function associated with matrix norms called the Schatten p-norms. These norms apply on matrix singular values. The proposed approach is illustrated upon scattering and coherency matrices on RADARSAT-2 PolSAR images over the Chamonix Mont-Blanc site. Several p values of Schatten p-norms are sur-veyed and their capabilities on filtering PolSAR images is provided in comparison with conventional strategies for fil-tering PolSAR data