Critical analysis of polarimetric PALSAR data for landcover classification

Polarimetric SAR provides a substantial source of information over conventional imaging radars. Fully polarimetric PALSAR data has been used for land cover classification. Co and Cross Polarization signatures for different classes are obtained in linear (h,v) and circular (l.r) basis. Different combinations from both basis have been extracted for the (principal component analysis) PCA. Supervised minimum distance classifier is used on principal components of various bases combinations for classification. HH-HV-VV feature set gives a better estimate of urban and vegetation class but water bodies are estimated with higher accuracy from LL-LR-RR feature set.