Classification of hybrid-pol data based on Euclidean distance between Stokes vectors

In this paper, a new classification technique for hybrid-pol SAR data based on Euclidean distance between Stokes vectors is introduced. The minimum Euclidean distance specifies the maximum similarity between two Stokes vectors which in turn indicates the maximum similarity between polarization behavior of corresponding backscattered waves. On the basis of this similarity, the backscattered wave from a scatterer is classified into three basic scattering mechanisms. We demonstrated that the proposed technique is able to correctly classify the three basic scattering mechanisms and performs better than existing hybrid-pol classification algorithms such as m - δ and m - χ.