FUZZY BASED MULTILEVEL MEDIAN FILTER FOR MIXED NOISE

Abstract

In this paper we have proposed fuzzy based multilevel median filter which is based on the fuzzy classical filter and the multilevel median filter. This filter is capableof reducing the mixed noise, which is the combination of salt & pepper noise, speckle noise and Gaussian noise with different levels of noise variances. The performance result of the proposed filter is found better than the conventional reduction filters such as median filter, Wiener filter and the asymmetrical triangular fuzzy filter with median center(ATMED), the asymmetrical triangular fuzzy filter with moving average center (ATMAV),the symmetrical triangular fuzzy filter with median center (TMED), and symmetrical triangular fuzzy filter with moving average center (TMAV). The comparison ofperformance is done by using MATLAB with numerical measurement of peak signal to noise ratio (PSNR).