Terrain Classification of Polarimetric SAR Images Based on Optimal Polarization Features

With the continuous progress of remote sensing technology, a large number of airborne and satellite-based polarimetric SAR have been applied in the field of remote sensing to provide decision basis for military intelligence, land use, and urban planning. Terrain classification of PolSAR image has become an important research direction in remote sensing image interpretation. In this paper, polarimetric SAR data of Flevoland area in the Netherlands are selected. First, the coherent speckle noise is suppressed by Refined Lee filtering, and 68-dimensional polarization features are obtained by the polarization target decomposition method. Then, based on the feature judgment factor and feature selection factor, the optimal polarization features are selected quantitatively and analyzed qualitatively with the help of Cloude-Pottier scatter plot. Finally, the selected optimal polarization feature combination is fed into the SVM classifier to realize terrain classification. The results show that the overall accuracy of classification in the Flevoland dataset reaches 95.7% using the method proposed in this paper, which effectively realizes the high-precision terrain classification of PolSAR images.