| 基于核主成分分析的虹膜识别方法 |
| |
| 引用本文： | 常卫东,王正华,刘完芳,鄢喜爱,张颜华. 基于核主成分分析的虹膜识别方法[J]. 兰州理工大学学报, 2007, 33(4): 86-89 |
| |
| 作者姓名： | 常卫东 王正华 刘完芳 鄢喜爱 张颜华 |
| |
| 作者单位： | 国防科学技术大学,计算机学院,湖南,长沙,410073;湖南公安高等专科学校,计算机系,湖南,长沙,410138;国防科学技术大学,计算机学院,湖南,长沙,410073;湖南公安高等专科学校,计算机系,湖南,长沙,410138;湖南大学,计算机与通讯学院,湖南,长沙,410083;湖南公安高等专科学校,计算机系,湖南,长沙,410138;湖南公安高等专科学校,计算机系,湖南,长沙,410138;中南大学,信息科学与工程学院,湖南,长沙,410083 |
| |
| 基金项目： | 国家重点自然科学基金(69933030)，公安部应用创新计划资助项目(2006YYCXHNST024)，湖南省公安厅2006年科研课题资助项目 |
| |
| 摘 要： | 阐述虹膜作为生物测定学特征用于身份识别具有得天独厚的优势,虹膜识别在场所或资源的安全控制等方面具有重要的应用价值.提出一种新的虹膜识别方法,该方法利用核主成分分析(KPCA)提取虹膜的纹理特征,通过竞争学习寻找其中最优的KPCA特征,形成虹膜编码,最后通过计算编码之间的方差倒数加权欧氏距离对虹膜进行识别.实验结果表明,该方法计算速度快,提取特征的效果好,对环境的适应性强,可用于实际的身份鉴别系统.
|
| 关 键 词： | 虹膜识别 核主成分分析 竞争学习 |
| 文章编号： | 1673-5196（2007）04-0086-04 |
| 修稿时间： | 2006-11-27 |
Iris recognition algorithm based on kernel principal component analysis |
| |
| CHANG Wei-dong,WANG Zheng-hua,LIU Wan-fang,YAN Xi-ai,ZHANG Yan-hua. Iris recognition algorithm based on kernel principal component analysis[J]. Journal of Lanzhou University of Technology, 2007, 33(4): 86-89 |
| |
| Authors: | CHANG Wei-dong WANG Zheng-hua LIU Wan-fang YAN Xi-ai ZHANG Yan-hua |
| |
| Affiliation: | 1. Institute of Computer Science, National University of Defense Technology, Changsha 410073, China; 2. Dept. of Computer Science, Hunan Public Security College, Changsha 410138, China; 3. College of Computer and Communication, Hunan University, Changsha 410083, China; 4. College of Information Science and Engineering, Central South University, Changsha 410083, China |
| |
| Abstract: | Having unique priority in biometric testing,iris recognition possesses an important widely application value in security control of places or resources.Iris recognition is composed of some processes such as iris localization,extraction of texture feature,and mode matching.A new iris recognition algorithm was proposed,where kernel principal component analysis(KPCA) was used to extract iris texture feature and a competitive learning mechanism was employed to choose the best feature from KPCA,then forming the codes of iris.Finally,the iris would be recognized by calculating Euclidean distance of covariance weighted reciprocal among the codes.The experimental results of iris recognition showed that this new approach exhibited high operational speed,better effect of feature extraction,and strong adaptability to environment.Therefore,it could be used in actual personal identification system. |
| |
| Keywords: | iris recognition kernel principal component analysis competitive learning |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
