Sparsity constraint SAR speckle reduction based on redundant multiscale ridgelet dictionary
Chengzhi Deng, Shengqian Wang · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2009
In this paper, we proposed a SAR speckle reduction method based on sparse and redundant representations over multiscale ridgelet dictionary. Firstly, the multiscale ridgelet function is proposed. And then based on it, the multiscale ridgelet dictionary is constructed, which can sparsely represent the SAR images. Finally, we propose a global image prior that forces sparsity over small patches in every location in the image. We define a maximum a-posteriori probability (MAP) estimator as the minimizer of a well-defined global penalty term. The speckle reduction leads to a simple iterated patch-by-patch sparse coding and averaging algorithm. The experimental results demonstrate that the proposed method performs better than several other existing methods in terms of quantitative performance as well as in term of visual quality of the images.