Salt and Pepper Noise Removal Algorithm Based on Neighbourhood Mean
Xianyi Qian · Journal of Nanjing University of Science and Technology · 2011
To improve the image effect,a filter algorithm for image noise removal suitable for salt and pepper noise is proposed based on the correlation of the neighbourhood.The pixels contaminated by salt and pepper noise are detected by using the maximum-minimum principle.Eight pixels near the contaminated one are divided into two classes according to their distances.The gray value of contaminated pixel is reconstructed by the mean of those uncontaminated pixels with near distance.The gray value of contaminated pixel is reconstructed by the mean of those uncontaminated pixels with far distance if all the pixels with near distance are contaminated.Computer simulation results show that the proposed algorithm with superior peak signal-to-noise ratio can restrain the noise and preserve the detailed information of images.