A symmetric PDF for model-based decomposition using Quad-POL data

摘要

A new general orientation probability density function based on tree symmetric structures is proposed to deal with the reflection symmetric scattering case for forest areas. The resultant general volume model can interpret the emergence of scattering reflection symmetry in the measured data. The new general volume model can be expanded to deal with the non-reflection scattering case. Compared with Arii's decomposition, the proposed decomposition has several advantages, lower remainder scattering power, more reasonable range of the orientation angle and the standard deviation. To verify the performance of the proposed decomposition, an L band airborne ESAR Quad-POL image is applied in the experiment.