摘要
目前广泛采用最大曲率方法实现指纹中心定位,但在指纹中心缺失情况下,该方法定位并不准确。为解决该问题,提出了一种基于相关参数模板匹配的指纹中心定位新方法,对FVC2004指纹数据库的测试表明,在指纹经过初步增强的基础上,借助最小距离判别,新方法能够对指纹中心准确定位,与最大曲率方法相比,在中心点缺失的情况下,有更高的鲁棒性。
To realize center point positioning of fingerprint, method of maximum curvature is widely used at present. But it is inaccurate when the center region of fingerprint has been lost or not clear. To solve that problem, a new method for center point positioning bases on template matching with parameters is presented. A test to FVC2004 fingerprint database indicates that the basis of elementary enha- ncement and least distance discrimination, the new method can position the center of f＇mgerprint with accuracy. Compared with method of maximum curvature, it has better robustness conditioning the center region lost.
出处
《计算机工程与设计》
CSCD
北大核心
2007年第9期2079-2080,2137,共3页
Computer Engineering and Design
基金
教育部博士点科研基金项目(20030532004)
关键词
指纹中心定位
参数模板
最大曲率法
指纹增强
最小距离判别
center point position of fingerprint
template with parameters
method of maximum curvature
fingerprint enhancement
least distance discrimination
