Recursive edge-preserving image filter based on a fuzzy model

In this paper, the problem of edge preserving smoothing in image processing is tackled using a fuzzy framework. The gray value differences between neighbour pixels are modelled as fuzzy variables and a design rule based on human experience and the nature of the filtering process gives rise to an effective filtering algorithm. Experimental results demonstrate the validity of the recursive fuzzy edge-preserving smoothing filter for both synthetic and natural images.