Four-Component Scattering Power Decomposition With Extended Volume Scattering Model

Abstract

In the three- or four-component decompositions, polarimetric scattering properties and corresponding physical scattering models play essential roles for power decomposition. This letter proposes an improved four-component scattering power decomposition method that employs a suitable volume scattering model for single- or double-bounce scattering in the polarimetric synthetic aperture radar image analysis. The cross-polarized <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">HV</i> component is created by both single-bounce object (such as vegetation) and double-bounce structures (such as oriented building blocks). It has been difficult to discriminate these two objects (vegetation against oriented buildings) in the decomposed images since the <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">HV</i> component is assigned to the volume scattering due to vegetation only. We propose to extend the volume scattering model suited for two physical scattering models. It is shown that a vegetation area and an oriented urban building area are well discriminated compared to those resulting from the implementation of the existing four-component scattering power decomposition.