An Algorithm for Removal of High Density Salt and Pepper Noise Based on Median Filtering
Zheng Jiang-yu · Journal of Anqing Teachers College · 2013
The process of median filtering which has been widely applied to remove salt and pepper noise is improved in this paper. The mean value of the non noise pixels centered at the current noise pixel in the 3 × 3 window is calculated and the gray of the noise pixel is replaced by the mean value. Then,the noise pixel is changed to non noise pixel. Because the noise pixel can immediately join to calculate the mean value of the back adjacent noise pixels,the removal of high density salt and pepper noise doesn't need to repeat and the size of windows keeps constant. The experimental results show that the proposed method behaves excellently for gray and color images which have different noise ratios. The proposed method can preserve the details of the original image well. Furthermore,the run time for the proposed method is quite short,which makes it possible for the real applications.