摘要
指纹匹配是指纹识别中的一个重要问题,直接影响着系统的正确率。目前普遍使用的匹配算法主要是基于节点的匹配算法。Ross提出了结合节点和纹理特征的指纹匹配方法,但Ross方法时间代价大。本文提出了结合节点和方向场的匹配算法,在用节点进行旋转和平移对齐的基础上,计算两个指纹方向场的一致性。实验结果表明:该算法可以比现有的基于节点的匹配算法更准确地区分两个指纹,同时比Ross方法更快。
Fingerprint matching is an important problem in fingerprint identification and directly decides the correctiveness rate of the system. Minutiae based matching is now pervasively used. The fingerprint matching approach using minutiae and texture features is presented by Ross. But Ross ＇ method is time-consuming. This paper presents an approach using minutiae and the orientation field. It first aligns two fingerprints based on their minutiae, and then computes the consistency of their orientation fields. Experiments show that this approach leads to substantial improvement in performance compared with the current minutiae-based approach and Ross＇method.
出处
《计算机工程与科学》
CSCD
2005年第9期54-56,共3页
Computer Engineering & Science
基金
国家自然科学基金资助项目(60373023)
关键词
指纹
匹配
节点
方向场
fingerprint
matching
minutia
orientation field
