Land cover classification of polarimetric SAR images for the Yellow River Delta based on support vector machine

Abstract

Land cover classification is one of the major measures for the ecological survey of river deltas. This paper focuses on the land cover classification of the RADARSA T-2 polarimetric synthetic aperture radar (PoISAR) images for the Yellow River Delta. As the effective utilization of polarimetric information is a key problem in PolSAR image classifications, we investigate the land cover classification for the Yellow River Delta based on the support vector machine (SVM). The study site, locates near the south of Bohai, in Shandong Province, China. The proposed method integrates several polarimetric target decompositions, PolSAR interferometry (PolInSAR), textural features derived from the gray-level co-occurrence matrix (GLCM), and the SVM. The traditional supervised Wishart classification is also performed for comparison. Experimental results validate the feasibility of the proposed method for land cover classification of the Yellow River Delta, Le., the overall accuracy reaches up to 90.91 %, while that for the method based on the Wishart distance is 85.01 %, which exhibits the superiority of the proposed method over the supervised Wishart method for polarimetric SAR images classification in the river estuary areas.