Class of nonlinear filters based on fuzzy membership functions

A novel filtering technique based on fuzzy logic is proposed for the enhancement of signals and images corrupted by a combination of different kinds of noise. The proposed nonlinear filters are based on fuzzy mapping known as membership functions. The proposed filters can selectively enhance or reduce various image details as membership functions can acquire varying shapes. These filters are referred to as fuzzy median filters. High performance of these filters is demonstrated in noise reduction of 1D and 2D test signals. Preliminary experimental results show that the proposed approach is effective and suitable for producing a unified treatment of different filtering tasks and can give better results than the classical filters