作 者: ;
机构地区: 许昌学院电气信息工程学院
出 处: 《计算机时代》 2011年第9期27-29,共3页
摘 要: 设计了指纹快速匹配算法。算法以分类时获取的奇异点为参考点对指纹进行定位,利用奇异点特殊邻近结构缩小细节点匹配规模,在可变界限盒的基础上设计合理匹配策略。该方法不仅可以提高计算速度,也能够适应指纹非线性形变的特点,增加算法鲁棒性。在FVC2000数据库上的测试结果显示,该算法在保证自动指纹识别系统识别准确性的同时,还使得指纹辨识速度有了显著的提高。 A fast fingerprint matching algorithm is designed. The algorithm locates fingerprints using the singular points obtained from classification as reference points, reduces the matching scale of minutiae by utilizing the special neighbor structure of the singular points, and designs reasonable matching strategy based on changeable sized bounding box. The algorithm can not only increase computing speed, but also adapt to the features of nonlinear deformation in fingerprints, which makes the algorithm more robust. The test results on FVC2000 databases show that this algorithm can speed up fingerprint identification with ensuring the accuracy of automated fingerprint identification system.
领 域: [自动化与计算机技术]