基于脊线校准的指纹细节点匹配算法

对自动指纹识别系统(AFIS, Automated Fingerprint Identification Systems)中居于重要地位的细节点匹配算法进行研究,对基于校准的细节匹配算法进行了修正.在细节点相关脊线信息的提取阶段,提取分叉点相关脊线的拓扑信息,并将其引入后续的校准和匹配算法.采用一种半可变的限界盒以适应指纹图像的非线性形变.引入一个新的参考量计算并比较脊线相似度,以更低的计算代价解决匹配中参考细节点对的选择问题,加快匹配算法的收敛速度.将脊线校准和细节点匹配两个阶段结合起来,采用一种新的更简单有效的方法进行细节匹配.测试结果表明,改进的匹配算法能够在使用较少资源的前提下,提高指纹匹配速度,改善系统的匹配性能.

Fingerprint minutia matching algorithm based on ridge alignment

Focus on the minutia matching algorithm which is important in automated fingerprint identification systems(AFIS).An improvement was made to the fingerprint alignment-based minutia matching algorithm.In the step of extracting minutia associated ridge information,ridge topologic information about bifurcation was extracted to use in the ridge alignment and minutia matching algorithm.The semi-changeable sized boundary box was used to make this algorithm robust to nonlinear deformation between fingerprint images.A new reference variable was introduced to compute and compare the similar degree of ridges;it can solve the problem of reference point pair selection with lower computational cost and faster convergent velocity of matching algorithm.Ridge alignment process and minutia matching algorithm were combined together,a simple and effective method was put forward in the alignment and matching algorithm.The experimental results show that the new method can greatly improve the matching speed and also develop the system's performance with less resource engaged.