Signal adaptive filter for noisy images based on fuzzy S- and T-norm operators

An extended version of a signal adaptive (SA) filter based on fuzzy S-norm and T-norm (S/T-norm) operators is introduced for image noise suppression. It is shown that by using Yager S/T-norm operators the proposed fuzzy filter provides a trade-off between the maximum likelihood (ML) and the signal adaptive Wiener (SAW) estimates of the true signal. By utilization of the fuzzy S/T-norm operators we provide coarser noise elimination when the variance of the desired signal relative to noise decreases. On the other hand, the proposed fuzzy filter adaptively leads to SAW filter when the variance of the desired signal relative to noise increases. Experimental results on several test images and for different noise variances show the feasibility of the proposed approach. The proposed method is also compared with other denoising filters by subjective and objective measures.