Robust Estimation Approach for Nonlocal-means Denoising Based On Structurally Similar Patches

Edge preserved smoothing techniques has gained importance for the purpose of image denoising. A good edge preserved filtering is given by nonlocal-means filter than any linear model based approaches.The contributions are in two fold. First, this paper explores a refined approach of nonlocal-means filter by using robust estimation function rather than the usual exponential function for its weight calculation. Here the filter output at each pixel is the weighted average of pixels with the surrounding neighborhoods using the chosen robust estimation function. Second, in order to speed up the computation, a new patch classification method is followed to eliminate the uncorrelated patches from the weighted averaging process. This patch classification approach compares favorably to existing techniques in respect of quality versus computational time. Validations using various test images have been analyzed and the results were compared with other known recent methods. There is reason to believe that this refined algorithm has some interesting and notable points.