基于直方图分解的去椒盐噪声算法

摘要

提出一种基于直方图分解的去椒盐噪声算法.根据椒盐噪声图像估计出原始图像直方图,对直方图进行分解得到高斯曲线,以曲线函数值作为权值对噪声图像进行加权均值滤波.并与中值滤波器、极值型中值滤波器以及模糊加权平均滤波器进行滤波效果对比.实验结果表明,该方法去噪性能较优.%This paper presents a denoising algorithm based on histogram decomposition.Histograms of original images are estimated from noisy images.Gauss curves are got by decomposing of estimated histograms and a weighted mean algorithm using the curve functions as weight is established.The performance of the method is compared with median filter, minmax filter and fuzzy filter wftu.Experimental results show that the performance of the new filter is better.