• 首页
  • 期刊简介
  • 编委会
  • 投稿指南
  • 收录情况
  • 杂志订阅
  • 联系我们
引用本文:白 列,蔡 芸,蒋 林.基于区域划分的非局部均值图像去噪算法的改进[J].软件工程,2023,26(5):1-5.【点击复制】
【打印本页】   【下载PDF全文】   【查看/发表评论】  【下载PDF阅读器】  
←前一篇|后一篇→ 过刊浏览
分享到: 微信 更多
基于区域划分的非局部均值图像去噪算法的改进
白 列,蔡 芸,蒋 林
(武汉科技大学冶金装备及其控制教育部重点实验室,湖北 武汉 430081)
2425191125@qq.com; 1caiyun@163.com; jlxyhjl@163.com
摘 要: 为改善非局部均值(Non-Local Means,NLM)算法的去噪性能,解决NLM算法参数分配以及去噪后图像边缘模糊等问题,对基于区域划分的非局部均值图像去噪算法进行了改进。通过Canny边缘检测算子和形态学膨胀处理对图像进行区域划分,对划分后的不同区域进行参数的调整,并对欧氏距离和权重函数进行改进,提升NLM算法的去噪性能,使去噪后的图像保留更多的细节纹理信息。实验结果表明,该算法相比于传统的NLM去噪算法、参数自适应的NLM算法以及基于转动惯量的改进权重函数的NLM算法,有着更好的峰值信噪比和结构相似度值。
关键词: 图像去噪;非局部均值;欧氏距离;权重函数
中图分类号: TP391    文献标识码: A
基金项目: 国家重点研发计划项目(2019YFB1310000).
Improvement of Non-Local Means Image Denoising Algorithm based on Region Division
BAI Lie, CAI Yun, JIANG Lin
(Key Laboratory of Metallurgical Equipment and Control Technology, Wuhan University of Science and Technology, Wuhan 430081, China )
2425191125@qq.com; 1caiyun@163.com; jlxyhjl@163.com
Abstract: In order to improve the denoising performance of the Non-Local Means (NLM) algorithm and solve the problem of parameter allocation and image edge blur after denoising, this paper proposes to improve the NLM image denoising algorithm based on region division. By Canny edge detection operator and morphological processing of dilation, image is divided into different regions where parameters are adjusted, and Euclidean distance and weight functions are improved, so that the denoising performance of the NLM algorithm is enhanced and the denoised image retains more detailed texture information. The experimental results show that the improved algorithm has better peak signal-to-noise ratio and structural similarity values, compared to traditional NLM denoising algorithms, parameter adaptive NLM algorithms, and improved weight function NLM algorithms based on rotational inertia.
Keywords: image denoising; non-local means; Euclidean distance; weight function


版权所有:软件工程杂志社
地址:辽宁省沈阳市浑南区新秀街2号 邮政编码:110179
电话:0411-84767887 传真:0411-84835089 Email:semagazine@neusoft.edu.cn
备案号:辽ICP备17007376号-1
技术支持:北京勤云科技发展有限公司

用微信扫一扫

用微信扫一扫