Image noise removal with detail preservation for random valued impulse noise

Abstract

Edge information is the most vital high frequency information of an image, filtering an image to reduce noise while keeping the image details preserved is one of the most important issues. In this paper we proposed a method in which double median filtering stage is used to preserve edge details, which enhance the quality of image. The algorithm is carried out in two stages; first stage is detection stage which detects noisy pixels, second stage is a filtering stage in which median is calculated twice. The goal of proposed algorithm is to remove high density noise in digital images while keeping edge details. The proposed method gives better image quality with high peak signal to noise ratio (PSNR) and reduced mean square error (MSE). Results of proposed algorithm have been analyzed in terms of visual and quantitative results.