An Efficient Threshold Based Mixed Noise Removal Technique

Abstract

Removing or reducing noises from image is very important task in image processing. This paper presents an efficient noise removal technique to restore original digital images corrupted by mixed noise. The proposed filtering technique consists of three steps: noisy pixel detection using fuzzy flag, mixed noise filtering step and calculating threshold value remove the pixel value with replacement conditions. Noises in this methodology are the combination of gaussian noise and salt and pepper noise. This methodology reduces the mixed noise without lossing edges sharpness and information. This methodology gives better results existing many fuzzy algorithms. The proposed technique shows better peak signal noise ratio result with thresholding replacement conditions. Hence, this mixed noise removal technique finds application in numerous segments of image process like digital tv, medical image process, camera, police work systems etc. Wiener filter is used for image enhancement.