SAR image data compression based on contrast modulation and warped wavelet Markov modeling

SAR image compression is very important in reducing the cost of data storage and transmission in relatively slow channels. In this paper, we propose a compression scheme driven by sharpness and contrast, rich texture information included in SAR images and speckle noise reduction within the wavelet framework. Firstly, we modulate and enhance the contrast of the original data in different scales using contrast pyramid, and then we use warped wavelet transform to capture geometric features of different scales. Finally Markov models are used to quantify how well a particular arrangement of wavelets fits our notion of edge structure. Compression performance is evaluated through subjective inspection, as well as objective performance measurements. Results clearly demonstrate the superiority of this new approach when compared to conventional SPIHT and JPEG systems.