光学学报, 2006, 26 (10): 1537, 网络出版: 2006-10-31
基于灰度曲面匹配的虹膜识别方法
Iris Identification Method Based on Gray Surface Matching
医用光学与生物技术 生物特征识别 虹膜识别 曲面匹配 方差 medical optics and biotechnology biometrics iris identification surface matching variance
摘要
目前的虹膜识别都是采用在图像上提取特征点,并将特征点编码为固定长度的特征数据用于匹配的方式。这种方式使虹膜识别系统易受攻击。为了进一步提高虹膜识别系统的安全性和识别速度,提出了一种基于灰度曲面匹配的虹膜识别方法。该方法抛弃了特征提取和编码等传统操作,在特征分析的基础上直接利用了灰度曲面匹配的思想,首先对两幅图像中的像素灰度做差,得到灰度差曲面,然后求出该灰度差曲面的方差。将此方差作为衡量两个虹膜特征曲面之间距离的依据,并据此判定两个虹膜是否来自同一只眼睛。在给定阈值为40的前提下,正确识别率为96.89%,识别时间为53.2 ms。实验结果证明,该方法识别准确率高,识别速度快。
Abstract
Nowadays, all of the iris identification methods are utilizing the strategy that extracts the feature points in the iris image firstly and then encodes these points into a certain length feature data for matching. But this strategy can make the iris identification system fragile to be attack. In order to improve the security and speed of iris identification systems, an iris identification method based on gray surface matching is proposed. This method abandons the traditional operations including the feature extracting and encoding, but uses the theory of gray surface matching based on feature analysis. Firstly, it computes the gray difference of two pixels from two different images and gets the gray difference surface. Secondly, it calculates the variance of the gray difference surface and considers this variance as the distance between two feature surfaces. At last, it decides whether these two irises are in the same eye or not according to the variance. When the distance threshold equals to 40, the accuracy is 96.89% and the speed is 53.2 ms. Experimental results demonstrate that this method has high accuracy and speed.
苑玮琦, 徐露, 林忠华. 基于灰度曲面匹配的虹膜识别方法[J]. 光学学报, 2006, 26(10): 1537. 苑玮琦, 徐露, 林忠华. Iris Identification Method Based on Gray Surface Matching[J]. Acta Optica Sinica, 2006, 26(10): 1537.