广义非局部均值算法的图像去噪

TP391.41; NLM (non-local means)滤波成为图像去噪关注的热点.该方法利用在图像中的结构特征冗余,对消除白噪声的效果较好,但对有色噪声效果不理想.对其作了改进,引入广义高斯分布模型以及马氏距离来取代欧氏距离,并且将其推广到图像序列的去噪领域中.结果表明,相较于NLM方法,该方法能够较好地抑制有色噪声,明显地改善了去除噪声效果,在保留图像纹理边缘的同时,有效地去除了图像中的噪声信息....

Full description

Saved in:
Bibliographic Details
Published in计算机应用研究 Vol. 32; no. 7; pp. 2218 - 2221
Main Author 郭红涛 王小伟 章勇勤
Format Journal Article
LanguageChinese
Published 华北水利水电大学软件学院,郑州,450045%郑州大学体育学院现代教育技术中心,郑州,450044%北京大学计算机科学技术研究所,北京,100080 2015
Subjects
信噪比
图像去噪
广义高斯模型
非局部均值
图像去噪
广义高斯模型
generalized Gaussian model
非局部均值
non-local means
信噪比
image denoising
signal to noise ratio
Online AccessGet full text

Cover

More Information
Summary:TP391.41; NLM (non-local means)滤波成为图像去噪关注的热点.该方法利用在图像中的结构特征冗余,对消除白噪声的效果较好,但对有色噪声效果不理想.对其作了改进,引入广义高斯分布模型以及马氏距离来取代欧氏距离,并且将其推广到图像序列的去噪领域中.结果表明,相较于NLM方法,该方法能够较好地抑制有色噪声,明显地改善了去除噪声效果,在保留图像纹理边缘的同时,有效地去除了图像中的噪声信息.
Bibliography:Image denoising is an important problem in computer vision and image processing. The non-local means （NLM） has received great attention in recent years, which makes use of the structural characteristics of image redundancy unlike conventional denoising algorithms based on local neighborhood. However, the NLM has good results to eliminate the white noise, whereas it is not effective for the colored noise. To solve this problem, this paper proposed a novel generalized nonlocal means denoising （GNLM） algorithm for noise removal of noisy images. With the introduction of generalized Gaussian distribution model,the proposed algorithm used the Mahalanobis distance instead of the Euclidean distance, and was also extended image sequence denoising. The experimental results show that the proposed algorithm improves the effect of the noise removal, and can do better on suppressing colored noise. The proposed algorithm can effectively eliminate image noise and significantly improve the image visual effect.
51-1196/TP
Guo Ho
ISSN:1001-3695
DOI:10.3969/j.issn.1001-3695.2015.07.073