Particle Swarm Optimization Based Parameter Adaptive SAR Image Denoising

This paper has put forward a novel SAR image denoising algorithm based on nonlocal means. In the traditional SAR image nonlocal means denoising algorithms, the patch similarity are measured by the accumulation of the pixel similarities. Considering that the similarity measured based on the norm of patches has got good denoising performance for the additive noise model, this paper has extended this idea to the multiplicative noise model for SAR image, and improved the Probabilistic Patch-Based(PPB) algorithm under the weighted maximum likelihood estimation framework. Since the parameters setting in PPB is complicated and it cannot adaptively get the best performance, this paper has proposed a particle swarm optimization based parameter adaptive nonlocal means algorithm for SAR image denoising. To check the performance of the proposed method, experiments compared with the canonical PPB method on real SAR image have been carried out. Experiments have demonstrated that the proposed method have a good performance on speckle reduction and details preservation.