【摘要】 概述并分析了常用的几种纹型分类方法,提出了一种基于方向射线和指纹图像空间域上的非严格对称属性的纹型分类算法。该算法对指纹中心点的定位要求低,纹型特征的维数低,有效减弱了常用方法中这两方面的计算偏差对分类结果的影响。以集合的方式定义出方向射线,进行指纹空间域特征提取,使得特征提取过程更直观,纹型分类策略更简单。最后实验证明了该算法的有效性。更多还原

【Abstract】 In this paper, several methods used to class fingerprint were introduced and analyzed. A novel spatial one was presented here, which made low request to fingerprint minutiae with low complexity by defining oriented radial and feature-indexing zone. The results from the experiment improve it effective.In this paper, several methods used to class fingerprint were introduced and analyzed. A fast fingerprint classification algorithm was presented, which was based on oriented radial and generalized nonsymmetrical features of fingerprint. The new algorithm made low request to fingerprint center location and the dimension of fingerprint features was also lower. The oriented radial was used to extract fingerprint spatial features. The process of extracting features and classification strategy were simple. At the bottom of this paper, results from the experiment improve it effective.更多还原