An efficient fuzzy inference system based approximated anisotropic diffusion for image de-noising

Fuzzy system is proven to be one of the well-effective approximation tool in soft computing techniques. The image denoising in modern multimedia system is strongly demanded issue which has been focussed in this work and an optimal solution with the help of fuzzy inference technique has been provided. An improved and approximated anisotropic diffusion scheme has been proposed by using fuzzy based diffusion coefficient functions. The anisotropic diffusion has been redefined by formulating the diffusion coefficients in terms of degrees of noisiness of each pixel which tends to sufficiently smooth the impulse noisy pixels along with preservation of edge pixels. The proposed fuzzy rule based diffusion coefficient is applied in basic Perona-Malik diffusion as well as selective advanced diffusion scheme and tested on various standard images at different noise densities. The proposed diffusion scheme based on fuzzy rule shows the effective results on images having impulsive noise densities upto 50%.