Abstract: Noise reduction is one of the most important parts of image processing.This paper analyzes the histogram of noise polluted images and presents a novel denoising algorithm based on the minimum error Bayes decision and the theory of meaning filtering.First the histogram of noise polluted images is counted and difference parameters are estimated.Then whether it is a noise or not to each pixel point in the image is decided.And the noise using meaning filtering is processed.The experiment results show that the algorithm presented in this paper is feasible and good.

Key words: image denoising, Bayes decision, meaning filtering, histogram

摘要： 去噪处理是图像处理中较为重要的环节。针对加噪后的图像的直方图进行分析，依据最小错误率贝叶斯决策和均值滤波理论，提出一种基于均值滤波和最小错误率贝叶斯决策的去噪方法。首先对加入噪声后的图像直方图进行统计，从中估计出服从分布的不同类别参数，对图像中每一像素点进行判断是否为噪声，对噪声点进行基于均值滤波的处理。通过试验，取得了良好的效果。

关键词: 图像去噪, 贝叶斯决策, 均值滤波, 直方图