Eigen-Decomposition-Based Four-Component Decomposition for PolSAR Data

This paper presents an alternative way of the decomposition scheme for polarimetric synthetic aperture radar (PolSAR) images. The proposed method intends to solve overestimation of volume scattering power for oriented urban areas with large orientation angle. In this paper, the difference of primary and secondary dominant target scattering orientation angle is proposed to show the possible cause of the volume scattering overestimation problem. The results of the eigen-decomposition method are introduced into the model-based decomposition. This paper applies a lookup table (LUT) to link eigen-decomposition method and model-based decomposition method together to solve the overestimation of volume scattering power problem. Based on the LUT, a new four-component decomposition method is proposed. L-band fully polarimetric ESAR and EMISAR data are used to validate the proposed method. The results of the experiments show that the proposed method has better performance than four-component decomposition method in oriented urban areas with large orientation angle. Besides, the volume model can also be directly determined from image. In this case, it is potential to solve the alternate volume scattering model problem in model-based decomposition.