摘要
指纹匹配是实现指纹识别的重要环节,针对如何提高识别率、减少算法复杂度,介绍了一种点模式的指纹匹配算法。其优点是:将匹配分为两步进行,减少了拒判时间;初匹配利用了细节点间的局部结构关系,既克服了图像的平移和旋转也得到了更精确的坐标校准参数;二次匹配阶段,整合多种判决条件作为识别的依据,提高了识别率。其实验结果表明,该算法复杂度低,识别率较高,有较强实用价值。
The purpose of this paper is to introduce a fingerprint matching approach based on minutiae . In this paper , we used the structral invariance of local minutiae to found local vectors, then integrate the best matching point and its neighbour points to get more accurate parameters of alignment adjustment, at last, we integrate multi-discriminant to complete fingerprint matching . The experiment data show that , our method have achieved a higher accuracy.
出处
《电子科技大学学报》
EI
CAS
CSCD
北大核心
2004年第2期154-157,共4页
Journal of University of Electronic Science and Technology of China
关键词
指纹识别
细节点
特征点集
局部向量
坐标校准
匹配
identification of fingerprint
minutiae
minutiae set
local vector
alignment adjustment
matching
