A new Six-Component Decomposition based on New Volume Scattering Models for PolSAR Image

Polarimetric target decomposition is a significant means for feature extraction of PolSAR image, which plays an important role in describing the scattering characteristics, physical structure, composition and other details of grounds. So in this paper, a new volume scattering model is proposed and applied to a new six-component decomposition, which decomposes the coherency matrix into surface scattering, double-bounce scattering, new volume scattering, cross scattering, oriented dipole scattering and compound dipole scattering. The proposed method can effectively improve the decomposition quality of PolSAR image, and its visualization presentation is more accurate than the existing decomposition methods. It is more robust in urban areas, suppress the problem of volume scattering overestimation, describe the complex environment in more detail, and obtain better scattering characteristics of man-made targets. The validity of the proposed method is demonstrated by the data of L-band fully polarimetric ESAR images of Oberpfaffenhofen area, Germany in this paper.