Using Cloude/Pottier Decomposition Method for Polarimetric Classification of RADARSAT-2 Satellite Image

The satellite polarimetric data does contain more information than the corresponding single polarization data. SAR polarimetry allows a discrimination of different types of scattering mechanisms because polarimetric signatures depend on the scattering process. The polarimetric classification is one of the most important applications of radar polarimetry in remote sensing. We present in this paper an unsupervised classification method that identifies target scattering characteristics. We used the entropy /alpha classification based on the Cloude/Pottier decomposition. Results are obtained using coherency matrix. We had a fully polarimetric image that is acquired from Quebec in Canada in the C band provided by the Canadian Centre for Remote Sensing CCRS.