Patch ordering based SAR image despeckling via SSC and wavelet thresholding

Abstract

In recent years, variety of techniques were developed in the field of SAR image despeckling, which avail to inhibit the Speckle in SAR Image. This paper proposes a patch ordering based SAR image despeckling approaches, which uses two transform domain filtering. The proposed approach consists of two stage filtering. In the first step i.e. coarse filtering, denoising is done by simultaneous Sparse Coding (SSC). The diminutive artifacts engendered by the coarse filtering can be removed by second stage of filtering i.e. refined filtering. In this step, filtered image is obtained by Wavelet Hard thresholding. Experimental results showed that the proposed system achieves good Structural similarity Index Measure (SSIM), Peak Signal to Noise Ratio (PSNR) values for despeckled images.