Compare Speckle Denoising Models Based on Total Variation and Filtering Methods

Abstract

Image denoising is among the most fundamental problems in image processing. A large range of methods covering various fields of mathematics are available for denoising an image. The initial denoising models are derived from energy minimization using nonlinear partial differential equations (PDEs). These methods are very fast and removes noise very well. The filtering based models have also been used for quite a long time where the denoising is done by smoothing operators. The most successful method among them was the nonlocal means proposed by Buades, Coll and Morel in 2005. Though the method is very accurate in removing noise, it is very slow and hence quite impractical. In 2008, Gilboa and Osher extended some known PDE and variational techniques in image processing to the nonlocal framework. The motivation behind this was to make any point interact with any other point in the image. Using nonlocal PDE operators, they proposed the nonlocal total variation method for Gaussian noise. Based on this, a nonlocal PDE based speckle denoising model has been developed. The model is faster than nonlocal means but still much slower than the total variation based models. Additionally in 2005 Mahmoudi and Sapiro improve the existing non local means model by using similar neighborhoods. In this paper, we compare the nonlocal tv based model with the faster nonlocal means model for removing speckle noise from images. We compare these models and try to find the better one.