基于加权超图和IOWA算子的图像椒盐噪声滤除算法

发布时间：2018-02-17 01:05

本文关键词： 椒盐噪声加权超图诱导有序加权平均(IOWA)算子图象复原　出处：《控制与决策》2017年04期 　论文类型：期刊论文

【摘要】：针对图像椒盐噪声,提出基于加权超图和诱导有序加权平均(IOWA)算子的椒盐噪声滤除算法.首先,用加权超图对图像进行表示,根据椒盐噪声为极值的特点,定义超图边的权值,该权值能够反映边内中心节点对应像素为噪声点的可能性,进而利用超图边的权值进行噪声检测;其次,构建IOWA算子对噪声点进行复原,并采用噪声检测与复原交替进行的方式完成图像的椒盐噪声滤除.仿真实验结果表明,所提出的算法不但可有效复原椒盐噪声,而且能保持原图像的轮廓等细节信息.
[Abstract]:In this paper, an algorithm for filtering salt and pepper noise based on weighted hypergraph and induced ordered weighted average IOWA-operator is proposed. Firstly, the image is represented by weighted hypergraph, and the weight of the edge of hypergraph is defined according to the extreme value of salt and pepper noise. The weight can reflect the possibility that the center node in the edge is a noise point, and then use the weight of the hypergraph edge to detect the noise. Secondly, the IOWA operator is constructed to restore the noise point. The simulation results show that the proposed algorithm can not only recover the salt and pepper noise effectively, but also keep the detailed information of the original image, such as the contour of the original image.
【作者单位】：安徽大学计算机科学与技术学院;安徽大学商学院;
【基金】：国家自然科学基金项目(61502003,71501002,61472002,61671018) 安徽省自然科学基金项目(1608085QF133,1508085QG149)
【分类号】：TP391.41

【相似文献】