Information Extraction from polarimetric SAR Images
No Thumbnail Available
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..
Supervisor: A. K. Mishra AND S. Dandapat
ELECTRONICS AND ELECTRICAL ENGINEERING