| 有效保持细节特征的图像椒盐噪声滤除方法 |
| |
| 引用本文： | 肖泉,丁兴号,王守觉,廖英豪,郭东辉.有效保持细节特征的图像椒盐噪声滤除方法[J].电子学报,2010,38(10):2273-2278. |
| |
| 作者姓名： | 肖泉 丁兴号 王守觉 廖英豪 郭东辉 |
| |
| 作者单位： | 厦门大学信息科学与技术学院电子工程系;中国科学院半导体研究所; |
| |
| 基金项目： | 国家自然科学基金，国家自然科学主任基金，福建省自然科学基金，厦门大学985二期信息创新平台，厦门市科技计划高校创新 |
| |
| 摘 要： | 论文结合几种去噪方法,提出一种统一的图像去噪模型.该模型通过一个统一的目标函数将图像去噪问题转化为最优化问题,目标函数的构造主要包括估计残差惩罚函数、局部权函数及正则化项三个方面.随后基于此模型提出一种新的去除椒盐噪声的非线性滤波方法,其中估计残差惩罚函数采用L1范数形式,局部权函数采用自适应高斯核函数,正则化项则利用图像的小波域稀疏性作为先验约束来构造.由于充分融合了图像的全局和局部统计特性,因而在抑制噪声的同时能够更好地保持图像边缘等细节特征,相关去噪实验结果证实了本文方法的有效性.
|
| 关 键 词： | 图像去噪 非线性滤波 自适应加权 细节保持 |
| 收稿时间： | 2008-01-13 |
A Novel Detail-Preserving Algorithm for Removing Salt-and-Pepper Noise |
| |
| XIAO Quan,DING Xing-hao,WANG Shou-jue,LIAO Ying-hao,GUO Dong-hui.A Novel Detail-Preserving Algorithm for Removing Salt-and-Pepper Noise[J].Acta Electronica Sinica,2010,38(10):2273-2278. |
| |
| Authors: | XIAO Quan DING Xing-hao WANG Shou-jue LIAO Ying-hao GUO Dong-hui |
| |
| Affiliation: | XIAO Quan1,DING Xing-hao1,WANG Shou-jue1,2,LIAO Ying-hao1,GUO Dong-hui1(1.School of Information Science and Technology,Xiamen University,Xiamen,Fujian 361005,2.Institute of Semiconductors of Chinese Academy of Sciences,Beijing 100083) |
| |
| Abstract: | A unified model for image denoising is proposed based on the analysis of several denoising methods.It by designing a unified objective function leads the image denoising into an optimization problem.The objective function builds up with three parts:the residual error penalty function,the local weighting function and the regularization term.Under this unified model,a novel nonlinear filtering algorithm for removing salt-and-pepper noise is subsequently proposed,in which the residual error penalty function in... |
| |
| Keywords: | image noising nonlinear filtering adaptive weighing detail-preserving |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《电子学报》浏览原始摘要信息 |
|
点击此处可从《电子学报》下载全文 |
|
