Comparison of SVM classifier and wish art classifier on L- band alos-palsar-2 data over metropolitan area

For every country, quantitative assessment of the Land Use and Land Cover (LULC) is essential for proper planning and for proper utilization of the resources nearby. Land cover change is related to global change due to its interaction with climate, ecosystem and from manmade activities. This paper focuses on Land cover classification of L band ALOS PALSAR Dual Pol data over the Metropolitan City Hyderabad. Longer wavelengths have more penetration capability, therefore, L band is opted for this study. The dataset is multilooked five looks in range and one look in azimuth direction, and speckle filtered with refined filter with window kernel size 3x3. In this study, we have compared the classification accuracy with two well know supervised classifiers VIZ Support Vector Machine (SVM) and Wishart Classifier. From this study, the classification accuracy for SVM and Wishart classifiers are almost similar i.e. 91.08% and 91.07%.