SAR Image Compression Using HVS Model
Wang Aili, Zhang Ye, Yanfeng Gu · 2006
Generally synthetic aperture radar (SAR) image compression methods based on wavelet transform remove statistic redundancy of image data and neglect visual redundancy. In view of this problem, a new image compression method based on human visual system (HVS) is proposed in this paper. First SAR image is decomposed by wavelet transform, then wavelet coefficients in different subbands are weighted by the peak of contrast sensitivity function (CSF) curve in wavelet domain, at last set partitioning in hierarchical trees (SPIHT) algorithm is used to code the weighted wavelet coefficients to form embedded bit stream. Compression results show that comparing with conventional SPIHT algorithm, the method proposed in this paper gets better subject visual quality at the same compression ratio with almost equivalent objective evaluation results