Land-cover Classification Based on SAR Data Using Superpixel and Cosine Similarity

Object-Based Image Analysis (OBIA) is one of the cutting edge image processing technologies attracting significant attention in remote sensing industry recent years and Synthetic Aperture Radar (SAR) possess a high potential for classification of earth surface. In this paper, a novel classification method, integrating OBIA and majority voting strategy, is proposed for land-cover classification based on SAR data. The main purposes of OBIA are delineating objects, extracting feature and measuring similarity by using superpixel segmentation technology and cosine similarity algorithm. The color histogram is utilized as feature in this study. Majority voting strategy provides an effective way for class determination. At last, the experimental result is presented and it demonstrates the proposed classification scheme is an appropriate method for land cover classification of SAR data.