Abstract: The algorithm of fingerprint matching based on graph is researched, which is composed of endings and bifurcations. The algorithm is stable according to the matching among the complete graphs composed of endings and bifurcations. The endings and bifurcations in the matching image are divided into different groups. The fingerprint identifies by using group and the template of images matching and it can reduce the complexity of computing effectively. The experimental results show that the proposed algorithm has a good quality and is very applicable even the image of fingerprint is distorted or has a low quality

Key words: Fingerprint recognition; Point pattern matching; Group; Complete graph

摘要： 研究了基于端点和分支点的点模式指纹匹配算法，通过对端点和分支点的完全图进行匹配，使其具有较好的稳定性。对匹配图像端点和分支点进行分组，通过每组的点和模板图像的细节点进行点模式匹配来实现指纹匹配，有效地降低了运算复杂度。实验结果表明，该方法在指纹产生非线性形变、指纹图像质量低等情况下有较好的性能，具有较强的实用性。

关键词: 指纹识别；点模式匹配；分组；完全图