Polarmetric SAR data compression using wavelet packets in a block coding scheme

Previously progress has been made in designing wavelet compression schemes for SAR image data. In this paper, we propose an improvement to the wavelet method based on a subdivision of the image into blocks, so that an optimal bit allocation can be made based on each block's statistics. Unlike single-channel SAR data, polarimetric data have HH, VV and HV channels of radar intensity images, plus the phase differences between them. We examine the information redundancy among these channels, and optimize the block coding scheme across the polarimetric channels and spatial blocks to improve compression performance. Experimental results using SIR-C data show significant improvement in coding efficiency.