首页 | 本学科首页   官方微博 | 高级检索  
     

基于人脸与虹膜的融合算法研究
引用本文:甘俊英,高建虎,李春芝. 基于人脸与虹膜的融合算法研究[J]. 计算机工程与设计, 2008, 29(3): 683-686
作者姓名:甘俊英  高建虎  李春芝
作者单位:1. 五邑大学,信息学院,广东,江门,529020;北京大学,视觉与听觉信息处理国家重点实验室,北京,100871
2. 五邑大学,信息学院,广东,江门,529020
基金项目:广东省自然科学基金 , 国家重点实验室基金
摘    要:对原始图像作对角变换,并在此基础上利用二维离散余弦变换(2DDCT)进行特征压缩,将人脸与虹膜特征相融合,利用二雏主元分析(2DPCA)分别从纵横双向进行特征提取,最后利用最小距离分类器进行识别.基于CAS-PEAL人脸数据库与CASIA虹膜数据库的实验结果表明,该方法降低了维数,有效地利用了分类特征信息,提高了正确识别率;证明了该算法的有效性,为多生物特征识别提供了一种新途径.

关 键 词:特征融合  图像对角变换  二维离散余弦变换  二维主元分析  多生物特征识别
文章编号:1000-7024(2008)03-0683-04
收稿时间:2007-03-12
修稿时间:2007-03-12

Research of fusion algorithm based on face and iris feature
GAN Jun-ying,GAO Jian-hu,LI Chun-zhi. Research of fusion algorithm based on face and iris feature[J]. Computer Engineering and Design, 2008, 29(3): 683-686
Authors:GAN Jun-ying  GAO Jian-hu  LI Chun-zhi
Abstract:Two-dimensional discrete cosine transform(2DDCT) is used in feature compression on the basis of image diagonal transform,and face and iris features are fused.Then 2DPCA is utilized in feature extraction from column and row direction respectively.Finally,nearest neighbor classifier is selected to perform recognition.Experimental results of CAS-PEAL face database and CASIA(Chinese Academy of Sciences,Institute of Automation) iris database show that the dimension is reduced,classified information is utilized ef-fectively,and the higher correct recognition rate is gained.In the meantime,the validity of the method is demonstrated,and a new approach is supplied for multimodal biometric identification.
Keywords:feature fusion   image diagonal transform   two-dimensional discrete cosine transform   two-dimensional principal component analysis   multimodal biometric recognition
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