A FIRE filter for detail-preserving smoothing of images corrupted by mixed noise

A new fuzzy inference ruled by else-action (FIRE) operator for the filtering of images corrupted by mixed noise is presented. By adopting a hierarchical structure, the proposed filter combines fuzzy rules for the cancellation of both impulse and uniform noise. Fuzzy sets and rules are specifically designed to yield an effective noise cancellation while presenting the image derails. The results of computer simulations show that the proposed approach is able to outperform a number of nonlinear methods in the literature.