Improved Classification of Polarimetric SAR Data Based on Four-component Scattering Model

In this, paper, we propose an improved classification algorithm which is based on the four-Component scattering model. Compared with the three-component model introduced by Freeman and Durden, the four-component scattering model introduces the "helix scattering" as its fourth component. Our algorithm emphasizes the existence of pixels with mixed scattering mechanism, and applies the result of decomposition as feature vector to initial merge and final iterative classifier instead of using the Wishart distance. We use L-band Pi-SAR images to demonstrate this new method. The experimental result verifies the effectiveness of this improved algorithm