Abstract: After investigating existing approaches to fingerprint classification, this paper introduces a novel method which is based on following the ridges of the fingerprint image. Other than utilizing core and delta points that are necessary in previous classification method, it first follows the ridges in the pattern area of a fingerprint image and gets a curve which can represent the area. After that some classification work can be do according to the curve’s parameters. Experiment results illustrate that the method can perform better.

摘要： 在深入研究现有指纹分类算法的基础上,提出了一种基于纹线跟踪的新颖指纹分类方法。不同于以往利用core点和delta点进行分类,该算法结合纹线跟踪的思想,在模式区进行纹线跟踪,得到模式区的跟踪曲线,然后再根据曲线参数进行分类,取得了较好的效果。

关键词: 纹线跟踪, 模式区, 曲线参数, 分类

CLC Number: 

• TP391.41