首页 | 本学科首页   官方微博 | 高级检索  
     

正则图像恢复中正则化算子选取的定性分析
引用本文:曾三友,康立山,丁立新. 正则图像恢复中正则化算子选取的定性分析[J]. 计算机工程与应用, 2003, 39(15): 42-43,210
作者姓名:曾三友  康立山  丁立新
作者单位:1. 株洲工学院计算机系,株洲,412008
2. 武汉大学软件工程国家重点实验室,武汉,430072;中国地质大学计算机学院,武汉,430074
3. 武汉大学软件工程国家重点实验室,武汉,430072
基金项目:国家自然科学基金资助项目(编号:60204001,70071042,60073043,60133010),武汉市青年晨光计划(编号:20025001002)
摘    要:对病态图像恢复中正则化算子的选取问题进行定性分析。以最小化正则解模糊误差为目标,利用泰勒级数定性地分析怎样选取正则化算子,得出结论:在信号的强度大于噪声强度的频带正则化算子应该是带阻的,而信号的强度小于噪声强度的频带正则化算子应该是带通的,通常情况下应选取低阻高通的正则化算子。实验结果表明了该结论的正确性。

关 键 词:正则化方法  图像恢复  正则化算子
文章编号:1002-8331-(2003)15-0042-02

A Qualitative Analysis of the Selection of Regularization Operator in Image Restoration
Zeng Sanyou Kang Lishan , Ding Lixin. A Qualitative Analysis of the Selection of Regularization Operator in Image Restoration[J]. Computer Engineering and Applications, 2003, 39(15): 42-43,210
Authors:Zeng Sanyou Kang Lishan    Ding Lixin
Affiliation:Zeng Sanyou 1 Kang Lishan 2,3 Ding Lixin 21
Abstract:The selection of regularization operator in image restoration is qualitatively analyzed in this paper.The focus is the minimization of the blurred error of regularized image.Taylor Expansion helps to choose regularization operator qualitatively,and yields a conclusion that in the frequency ranges of spectrum where the strength of image is greater than that of noise the regularization operator should be band-stop and in the frequency ranges of spectrum where the strength of image is smaller than that of noise the regularization operator should be band -pass and in the usual situation the regularization operator should be low-stop and high-pass.Numerical results illustrate that the conclusion from the qualitative analysis is right.
Keywords:Regularization method  Image restoration  Regularization operator  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号