An extended three-component scattering power decomposition for polarimetric SAR data

Abstract

This paper presents an extended three-component scattering power decomposition method that employs the co-polarized correlation coefficients in linear and circular polarization bases to analyze the dominant scattering mechanism that gives rise to the observed backscatter in polarimetric synthetic aperture radar (SAR) data. For the targets whose dominant scattering mechanism can be explicitly confirmed, the best fit dominating contribution is first computed using the general volume, surface or double bounce scattering models, and subsequently the residual components from the remainder power. By applying the new decomposition scheme to fully polarimetric SAR images, it is shown that the new method not only inherits the advantages of the classic three-component scattering power decomposition but also gives prominence to the targets characterized by strong volume, surface and double bounce scatterings.