A Mixed Noise Filter Acceleration Algorithm Based on Mean Value and Variance Similarity

A mixed noise filter acceleration algorithm base on mean value and variance similarity is proposed, Which is a improvement of existing better filter algorithm MNF in the present. The algorithm is to Classify of mean value and variance similarity is a local neighborhood of a pixel, the around pixels are divided into two types: similar or not similar. Similar pixels participate in filter algorithm, on the contrary do not. Thus, the algorithm effectively improves the filter speed. Simulation experimental results show that compared with MNF the algorithm greatly increased the computing speed, and the filter effect is not reduced.