Abstract: “Image denoising is considered as an effective initial processing step in different sophisticated imaging applications. Over the years, numerous studies have been conducted filtering of different kinds of noise. The Block Matching with 3D filtering added a new dimension to the study of denoising techniques. The main motive of this work is to establish a novel denoising method for multiplicative noise (speckle), with the ideology of the sparse representation in filtering via Maximum a Posteriori (MAP) estimator approach that is used in 3D array performed by grouping similar patches. Experimental results justified a good performance of novel framework that appears to demonstrate better denoising performance against state-of-art algorithms according objective criteria (PSNR, SSIM, EPI) values as well as subjective visual perception.”

Keywords: Denoising, 3D filtering, MAP estimator