基于方向的指纹奇异点提取  

Singular Point Detection of Fingerprint Based on Direction

作　　者：丁晋俊[1] 孙乐昌[1] 

机构地区：[1]解放军电子工程学院网络工程系,安徽合肥230037

出　　处：《计算机技术与发展》2007年第2期109-110,114,共3页Computer Technology and Development

摘　　要：准确、可靠地检测指纹奇异点(核心点和三角点)对于指纹的分类和匹配有重要的意义。针对指纹图像奇异点提取中准确判断和精确定位的难题,介绍了一种比较好的奇异点检测算法。根据奇异点的性质,利用Poincare Index方法求出核心点和三角点。根据相关规则,清除虚假奇异点。实验结果证明该方法能够从指纹图像中较精确、可靠地提取出奇异点。用该方法对不同质量的指纹图像进行实验,并与其他方法进行比较,结果表明该方法更加有效、可靠,具有很好的鲁棒性。It is very important to detect singular point （core and delta） accurately and reliably for dassification and matching of finger- print. In this paper, introduce a good method for singularity detection in fingerprint to improve accuracy of the position and reliability of the singular point. Core and delta are detected by the Poincare Index method. The falsehood singular points are eliminated in basis of correlation rule. At last,experiment results show that the method detect accurately and reliably the singular points. Experimental results on fingerprint images and comparison with other methods show less computational complexity and more robustness.

关 键 词：指纹图像 方向图 奇异点 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]