SAR image compression combining with denoising based on multiwavelet spatial-orientated tree

Synthetic aperture radar (SAR) images compression is very important in reducing the burden of data storage and transmission in relatively slow channels. The relatively new transform of multiwavelets can possess desirable features simultaneously such as short support, orthogonality and symmetry, while scalar wavelets cannot. And the spatialorientated tree (SOT) is an efficient data structure to investigate the spatial similarity correlations among multiwavelet coefficients at different resolutions. Thus a compression scheme combing speckle noise reduction based on SOT within the multiwavelet framework is proposed. At first we utilized the SOT to normalize the multiwavelet coefficients in high frequency bands and removed the signal-dependence of the speckle noise. Then multiwavelet coefficients were coded to form embedded bit stream by modified set partitioning in hierarchical trees (SPIHT) algorithm. Experimental results show this coding method achieves favorable peak signal to noise ratio (PSNR) and superior speckle noise reduction performances.