Similarity hypergraph representation for impulsive noise reduction

In this paper, a new approach to the problem of impulsive noise reduction in image is presented. First, an image neighborhood hypergraph representation using a similarity measure is computed. Next, a detection procedure based on hypergraph properties is used to classify hyperedges either as noisy, or clean data. Then we apply a nonlinear filter to noisy detected pixels. The results show that the proposed method outperforms most of the basic algorithms for the reduction of impulsive noise.