An adaptive non-local means algorithm for image denoising via pixel region growing and merging

In this paper, we propose a new adaptive non-local means algorithm for image denoising based on the region via pixel region growing and merging. For a seed pixel, by employing the pixel region growing and merging technique to the search regions, it lets the search regions cluster under the similar gray and structure adaptively and merge the related neighborhoods. Experiment results demonstrate that this method is efficient and robust not only for Gaussian white noise, but also for salts and peppers noise. Compared with ONL means and NLR means, the proposed method can not only accelerate the computation, but also have a high quality.