一种新的基于完整相位分析的虹膜分类算法  

Novel iris classification algorithm based on complete phase analysis

作　　者：岳学东[1] 刘洋[2,3] 

机构地区：[1]郑州大学物理工程学院,郑州450052 [2]中国科学院西安光学精密机械研究所,西安710119 [3]中国科学院研究生院,北京100039

出　　处：《计算机工程与应用》2008年第23期201-202,211,共3页Computer Engineering and Applications

摘　　要：将预处理后的虹膜图像分成64幅子块,并分别进行特征提取,然后按顺序提取每幅子图像完整的相位信息,建立相位特征向量,最后通过对特征向量进行内积运算获得虹膜样本之间的归一化分类距离。实验结果表明,该算法思路简洁,识别效果好,为虹膜分类研究提供了一种新的参考途径。The preprocessed iris image is divided into 64 sub-blocks,and the iris features are extracted respectively.The complete phase information of the features is extracted to establish the feature vector of iris,and then the normalized classification distance of iris samples is obtained by plot operation.The experiment shows that the idea of the algorithm is concise and reasonable,and its recognition effect is better.What＇s more,it also offers a new way for iris classification research.

关 键 词：特征提取 相位相关 虹膜识别 分类距离 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]