Removal of random-valued impulsive noise from corrupted images

This paper presents a novel method for the suppression of random-valued impulsive noise from corrupted images. The proposed method is composed of an efficient noise detector and a pixel-restoration operator. The noise detector has been used to discriminate the uncorrupted pixels from the corrupted pixels. The noise-free intensity values of the corrupted pixels have been computed by using triangle-based linear interpolation and the values of tuning parameters of the proposed method have been optimized with differential evolution algorithm. Extensive simulation experiments indicate that the proposed method significantly outperforms all of the comparison methods mentioned in this paper. The success of the proposed method over comparison methods is due to its excellent detail preservation performance independent from the level of noise density.