Texture-Preserving Image Deblurring
Chen Fen, Xiaojun Huang, Wufan Chen · IEEE Signal Processing Letters · 2010
In this letter, a two-step texture-preserving image deblurring method is proposed. This method restores the blurred image in the frequency domain to obtain a noisy result with minimal loss of signal components, followed by a modified non-local means (NLM) filter to attenuate the leaked colored noise. This modified NLM filter computes the weights not only using the self-similarity of the neighbors, but also on spatial distance from the pixel in question. Experimental results demonstrate that the proposed algorithm can effectively obtain a visually pleasant texture-preserving result with compared to the state-of-the-art total variation method.