Selection Of Weighting Coefficients For Adaptive Weighted Mean Filter

Digital image restoration is to use the existing image information to automatically recover the original image as much as possible.The damage uniformly distributed image is a special kind of image,whose missing pixels spread evenly over the whole image,like the salt-pepper noise,thus the image could be repaired using the methods of salt-pepper noise filtering.In this paper several methods for filtering the salt-pepper noise are first introduced.Then discussion is focused on the issue of correlation among the adjacent pixels in a gray image,as the basis for the selection of the weighting coefficients used in an adaptive weighted mean restoration.Finally the weighting coefficient is determined after an orthogonal decomposition of the original image histogram using the properties of Gaussian distribution.The experimental results show that this improved method is simple and effective.