On the Mixed Scattering Mechanism Anal- Ysis of Model-based Decomposition for Po- Larimetric Sar Data

This paper introduces a simple but effective scattering mechanism identification scheme for analyzing mixed scattering mechanisms obtained by model-based decomposition. Using the normalized scattering vector, each pixel is represented by a point in a standard 2-simplex in R. Seven scattering category centers are represented by the three vertices, the three midpoints of sides and the centroid of the 2-simplex. The scattering category partitioning problem is then solved by minimizing the Euclidean distance between the image pixels and these category centers. The proposed scattering mechanism identification scheme is finally used for data analyzing and unsupervised classification. Experiments on AIRSAR and E-SAR Lband PolSAR images demonstrate the effectiveness of the proposed method.