摘要
为了满足嵌入式指纹识别系统的实时性需求,论文提出了一种基于ARM9内核的指纹匹配算法。该算法通过构建特征点的三角结构向量,消除了指纹图像旋转、平移的影响,提高了系统的鲁棒性。同时算法采用两个等级进行指纹匹配,缩短了匹配时间,提高了匹配效率。通过在ARM920T主建的平台上进行识别测试,本算法完全可以满足嵌入式系统的要求。
In order to satisfy the efficiency of embedded fingerprint recognition system,a fingerprint matching algorithm based on ARM9 is designed in this paper.Vector triangles of minutia is been constructed to make rotation or scale change invariant,the robustness of system is been improved.To obtain high operation rate,fingerprint has been matched by two steps.The test results on ARM920T show that this algorithm can satisfy the requirement of real-time system.
出处
《计算机与数字工程》
2013年第5期812-815,共4页
Computer & Digital Engineering
基金
天津科技大学科学研究基金项目"基于大规模指纹库的指纹算法研究"资助
关键词
指纹识别系统
预处理
特征点
矢量三角形
可变限界盒
AFIS
previous process
fingerprint feature
vector triangle
changeable sized boundary box
