Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (32): 156-158.DOI: 10.3778/j.issn.1002-8331.2010.32.043

• 图形、图像、模式识别 • Previous Articles     Next Articles

Zero-crossing detection and circle coding for iris recognition

YIN Yong,LI Rong-gang,SHI Zhen-ning   

  1. College of Communication Engineering,Chongqing University,Chongqing 400044,China
  • Received:2010-02-23 Revised:2010-05-20 Online:2010-11-11 Published:2010-11-11
  • Contact: YIN Yong

虹膜识别的过零检测循环编码方法

印 勇,李荣岗,史振宁   

  1. 重庆大学 通信工程学院,重庆 400044
  • 通讯作者: 印 勇

Abstract: The algorithm of extracting and coding the texture of iris is studied.The iris image must be preprocessed firstly to get the image of the same size.Then deal the image with Gaussian filter and wavelet transform zero-crossing,deal the series of functions getting from the iris image with secondary circle coding to get the eigenvalue of the iris image.Distinguish the different iris image eigenvalue by Euclidean distance last.Experimental results indicate that the mean recognition rate is about 96.3%.The algorithm is simple,it can solve the problem of the round inflexibility in iris recognition,the results of the experiments are effective.

摘要: 研究了虹膜特征提取和编码识别算法。首先对虹膜图像进行预处理获得归一化的虹膜图像,然后对归一化后的虹膜图像进行高斯平滑滤波和小波过零检测,对得到的特征矩阵通过二次循环编码以得到虹膜的特征值序列,最后通过改进的欧氏距离对虹膜进行判别。实验结果表明,识别率可以达到96.3%。该算法运算简单,能够解决虹膜识别中的旋转不变性问题,得到了较好的识别效果。

CLC Number: