消除指纹纹线中的伪特征

Eliminating False Feature of the Ridge Lines

文中介绍了自动指纹识别系统中,灰度指纹图像经过预处理过程得到细化二值图像;其中可能含有大量的伪特征,对匹配操作造成不良影响,导致识别率下降。为了提取出准确的、真实的特征点,选取自动阈值、对短线、断线、小桥等,并利用改进的方法去除,对区域边界的伪特征、三角和岛屿采用新的方法去除。

In AFIS （Automated Fingerprint Identification System）, the gray fingerprint image is transformed into a thinned image in which there are a mass of false features that affect the following verification and reduce the identification rate of the system. In this paper, an elimination algorithm of false features in fingerprint image is applied for getting exact and true features. Automatic threshold is used before the elimination false features. A few improved meth- ods and some new ways are achieved for the purpose of different structures of false features.