Analysis of Spatial Statistics in Polarimetric SAR Data Using Wavelet Signatures

The Wavelet Polarization Signature (WASP) is a formalism, recently introduced in the literature, for analysing in SAR imagery the dependences of texture measures afforded by wavelet frames on scale and polarization state. In this paper experiments are reported where WASP analysis is applied for the spatial statistics characterization of targets of interest in land and marine applications. Results indicate that the effects related to the behaviour of spatial statistics under a change in polarization state, which were predicted theoretically, do indeed occur and could be potentially useful in segmentation and target detection problems exploiting polarimetric texture diversity.