基于局部特征结构的指纹匹配算法研究

针对实际应用中指纹识别系统对畸变图像识别率较低的情况,提出了一种点模式指纹匹配的新方法,利用细节点方向的分布特点,构造指纹图像中具有平移旋转不变性的局部特征结构,然后根据模板指纹和输入指纹局部特征结构的相似度寻找候选参考点对,将指纹图像进行对齐,最后进行特征点集的全局匹配,由结果分数来判断两个指纹是否匹配,给出了算法实现的全过程,实验结果表明该算法鲁棒性好,正确识别率较高,是一种实用的指纹识别技术.

Research of Fingerprint Matching Algorithm Based on Similarity of Local Characteristic Structure

At present,most fingerprint verification systems can not get a satisfying result when working on some low quality fingerprint images.The purpose of this paper is to introduce a novel fingerprint matching algorithm based on minutiae.Firstly,the local characteristic structure with invariants under shift and rotation is constructed.Then the similarity of two local characteristic structure(one from template fingerprint,the other from input fingerprint)is calculated to choose a best pair of reference point.At last the result is given by global matc- hing.Through experiments,the accuracy of this algorithm is proved effective,even to low quality images,and it is expected to be a conven- ient technique.