Multiresolution Model for Image Denoising Based on Total Least Squares

In this paper, we propose a novel method for removing noise from images corrupted with mixed noise, which is based on total least square algorithm. The method of multiresolution was applied to our arithmetic with more concentrational strategy, which has improved the efficiency of denoised arithmetic for TLS image as well as shortened the runtime, while ensuring the quality of denosing, especially the point, line, and edge regions. The simulation demonstrates that our algorithm proposed has a better denoised effect compared with other traditional denoising technique.