Information extraction from polarimetric sar images

Abstract

This thesis deals with extraction of information from polarimetric Pol syn thetic aperture Radar SAR images The complete thesis work can be pre sented under three major areas The first major area of work comprises of information extraction from polarimetric SAR PolSAR images In this we propose and develop three new unsupervised algorithms for landcover mapping and crop classification using PolSAR images The second major area of work deals with extraction of information from the images of the recently introduced hybrid PolSAR architecture At first to realize the potential of this new po larimetric configuration a comparison of hybrid Pol configuration with the full polarimetric SAR configuration based on the information content is carried out Secondly we propose three new approaches for the analysis of the hybrid Pol data to extract information from its images The third major area of work deals with PolSAR image enhancement techniques More precisely this deals with sidelobe noise suppression in PolSAR images using non linear apodization techniques We also validate that better information can be extracted from the PolSAR images if sidelobe noise is removed from its images newline