An Algorithm for Image Denoising Based on Mixed Filter
Yanchun Wang, Dequn Liang, Heng Ma, Yan Wang · 2006
Generally, images can be corrupted by different characteristic noises simultaneously, we cannot obtain satisfactory filtering result if only using single filter such as average filter or median filter. Therefore, in this paper a new mixed filter algorithm is proposed for filtering the image corrupted by difference noises. Firstly, we construct adaptive structure using neighborhood contrast measure; secondly, divide the image into smoothness, edge and unconfirmed regions based on the adaptive structure; then, adopt corresponding filter for different regions. The algorithm does not need a priori knowledge of images and noises. We perform experiments on the image corrupted by Gaussian and impulse noises, by using average filter with maximization and minimization for the smoothness region, unidirectional median filter for the edge region and median filter for the indefinite region. The experiments show that the proposed algorithm is feasible and efficient