首页 | 本学科首页   官方微博 | 高级检索  
     

非下采样Contourlet域GCV准则SAR图像去噪*
引用本文:杨晓慧,焦李成.非下采样Contourlet域GCV准则SAR图像去噪*[J].计算机应用研究,2009,26(9):3542-3544.
作者姓名:杨晓慧  焦李成
作者单位:1. 河南大学,数学与信息科学学院,应用数学研究所,河南,开封,475004
2. 西安电子科技大学,智能信息处理研究所,智能感知图像理解教育部重点实验室,西安,710071
基金项目:国家自然科学基金资助项目(60802061);河南省教育厅自然科学基金资助项目(2008B510001)
摘    要:首先对SAR图像作非下采样Contourlet分解,充分考虑其系数统计特性,给出非下采样Contourlet域GCV准则,对每个分解层的各个子带作多层阈值估计和软阈值收缩处理,进而详细探讨分解层数和方向分解数对NSCT性能的影响。实验结果表明,该方法从视觉效果和客观衡量指标两方面都取得了比较理想的效果。

关 键 词:合成孔径雷达    图像去噪    非下采样Contourlet变换    广义交叉验证

SAR image denoising based on nonsubsampled Ministry domain GCV
YANG Xiao-hui,JIAO Li-cheng.SAR image denoising based on nonsubsampled Ministry domain GCV[J].Application Research of Computers,2009,26(9):3542-3544.
Authors:YANG Xiao-hui  JIAO Li-cheng
Affiliation:(1. Institute of Applied Mathematics, School of Mathematics & Information Sciences, Henan University, Kaifeng Henan 475004, China;2. Intelligent Perception & Image Understanding Key Laboratory of Ministry of Eduction of China, Institute of Intelligent Inf
Abstract:Firstly, decomposed synthetic aperture radar (SAR) image by the NSCT and statisticed the coefficients. And then established the generalized cross-validation (GCV) criterion in NSCT domain to obtain the multithresholds in each subbands of every decomposition leveles adaptively. The soft shrinkage was finally fufill the denoising. Finally, disscussed the influences to NSCT of decomposition levels and direction decomposition values. Experiments show that there are desiresresults both in terms of image visual fidelity and in terms of objective indexes.
Keywords:SAR  image denoising  nonsubsampled Contourlet transform(NSCT)  generalized crossed validity (GCV)
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《计算机应用研究》浏览原始摘要信息
点击此处可从《计算机应用研究》下载全文
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