Noise filter of multi-value images based on fuzzy Renyi entropy

Aim at the problem of noise-reduction for multi-value images, we propose a method for noise filter of multi-value images based on fuzzy Renyi entropy. The method builds up a relevant physical model according to the characteristics of degenerative multi-value images, and constructs a type of fuzzy entropy which can reflect the actual circumstance of images, and finally carries on the picture instauration using principle of minimum entropy. The experiments conducted on simulator show the superiority of noise-reduction through compared with median filter and average filter.