ABSTRACT

Reducing noises from images is a very wide research area in the domain of image processing. An efficient noise removal technique based on median filter to restore gray scale images corrupted by mixed noise, by preserving image details has been presented in this paper. This filtering method consists of two steps: the noisy pel detection and the mixed noise filtering. Noises addressed in this method are mixed impulse noise and Gaussian noise. The technique reduces mixed noises significantly without reducing on edge sharpness. Experimental results show that the proposed method outperforms the classic median and mean filtering algorithms, along with stabilizing the tradeoff between noise removal and image detail protection. So, this technique find relevance in various image processing systems including digital television, medical image processing systems, digital camera, surveillance systems etc.