IMPROVED ADAPTIVE FUZZY PUNCTUAL KRIGING FILTER FOR IMAGE RESTORATION

In this paper, we present an intelligent image restoration approach by combining the geostatistical interpolation technique of punctual kriging, fuzzy logic with type II and fuzzy smoothing based approaches. Images degraded with Gaussian white noise are restored by first utilizing fuzzy logic for selecting pixels that needs kriging. Fuzzy type-II inference system is employed. Type-II fuzzy set has been used to generate fuzzy map for detection of noisy pixels. This fuzzy map of noisy pixels enhances performance of the proposed technique in terms of image restoration and as well as in terms of computational cost. Local neighborhood information is used to ensure noise free pixels in 3 x 3 window to estimate the noisy data. The concept of punctual kriging is then used to estimate the intensity of a noisy pixel. Image restoration performance based comparison has been made against adaptive Weiner filter and existing fuzzy kriging approaches. Experimental results using 450 images and different image quality measures show effectiveness of the proposed approach with fuzzy type-II for detecting noisy pixels and utilization of local information in conjunction with kriging based estimation.