Spatially adaptive image deblurring based on nonlocal means

The deconvolution of blurred and noisy images is an ill-posed inverse problem, which can be regularized under the Bayesian framework by introducing an appropriate image prior. In this paper, inspired by the state-of-art nonlocal means(NLM) denoising technique which exploits the similarity of the image patches, we construct an inhomogeneous and anisotropic image prior under the Markov random field theory, and thus propose a spatially adaptive deblurring method, called NLM-based deblurring method (NLMD). NLMD is capable of preserving the image's non-smooth structures, such as edges, textures. The experiments illustrate NLMD's potential and demonstrate that NLMD performs competitively compared to the best existing state-of-art debluring methods.