A Novel Polarimetric SAR Data Decomposition Method with Generalized Volume Scattering Model

Abstract

The Polarimetric target decomposition method based on scattering model is a hotspot, but there is a problem that decomposition results are inconsistent with actual scattering of ground objects. In order to solve the problem, a new polarimetric SAR data decomposition method based on generalized volume scattering model is proposed in this paper. Matrix rotation theory on basis of the dominant scatterer is used in the method, which removes cross-polarized scattering power generated by the coupling between orthogonal states of polarization. On this basis, a suitable volume scattering model is selected, which ultimately makes decomposition results are closer to scattering characteristics of ground objects and reduces the number of negative power pixels. Experiments show that the results of novel decomposition method reflect the scattering characteristics of real objects more accurately.