RANDOM-VALUED IMPULSE NOISE REMOVAL USING FUZZY LOGIC

In this paper, a fuzzy based random-valued impulse noise filtering technique is proposed. The proposed method is based on intelligent noise detection, intelligent image filtering and a detail preservation process. The noise detection method used in this paper is based on fuzzy gradients to detect the noisy pixels intelligently by differentiating them from the edge pixels. The detected noisy pixels are then processed by fuzzy based filtering process. Fuzzy based filtering process uses trapezoidal shaped fuzzy membership functions constructed through fuzzy set construction algorithm to remove random valued impulse noise from digital gray scale images. In order to preserve image details while filtering the noisy edge pixels, a fuzzy based detail preservation process has been introduced in the proposed method. The main contribution of the proposed technique includes 1) marvelous detail preservation at both noise detection and noise filtering level 2) removing the random-valued as well as mixed (salt & pepper and random-valued) impulse noise. Based on peak-signal-to-noise ratio and subjective quality measure, we have found experimentally that the proposed technique is superior to the state-of-the-art techniques for removing random-valued as well as mixed impulse noise.