J4

• 计算机科学 • 上一篇    下一篇

一种新的基于多小波变换的虹膜识别算法

魏连鑫1,2, 朱伟民3, 吴登峰3,4, 徐涛3   

  1. 1. 吉林大学 数学研究所, 长春 130012; 2. 上海理工大学 理学院, 上海 200093;3. 吉林大学 机械科学与工程学院, 长春 130022; 4. 长春师范学院 计算机科学与技术系, 长春 130032
  • 收稿日期:2006-04-07 修回日期:1900-01-01 出版日期:2007-01-26 发布日期:2007-01-26
  • 通讯作者: 徐涛

A New Iris Identification Method Based on Multiwavelet Transform

WEI Lianxin1,2, ZHU Weimin3, WU Dengfeng3,4, XU Tao3   

  1. 1. Institute of Mathematics, Jilin University, Changchun 130012, China; 2. College of Sciences, University of Shanghai for Science and Technology, Shanghai 200093, China;3. College of Mechanical Science and Engineering, Jilin University, Changchun 130022, China;4. Department of Computer Science and Technology, Changchun Normal University, Changchun 130032, China
  • Received:2006-04-07 Revised:1900-01-01 Online:2007-01-26 Published:2007-01-26
  • Contact: XU Tao

摘要: 提出一种基于多小波变换的新的虹膜识别算法. 利用CL多小波优良的数学性质, 虹膜纹理特征能够被准确完整地表达. 将分解后子带作为特征编码, 利用空间向量的方向夹角余弦作为匹配方法, 有效克服了缩放、 平移、 光照等影响, 达到了较好的识别效果.

关键词: 虹膜识别, 多小波变换, 纹理特征

Abstract: This paper is proposed a new method for iris identification based on multiwavelet transform. By means of the mathematical properties of CL multiwavelet, the iris texture feature can be presented accurately and primely. The subbands are coded and the information of direction cosine is used for iris matching, by means of which this method overcomes zoom, shift and illuminance and obtains approving effect.

Key words: iris identification, multiwavelet transform, texture feature

中图分类号: 

  • TP752