基于分类器的身份证号码识别研究

Identity card identification based on classifier

目前,许多场合下需要进行身份证自动识别,因此准确识别身份证号码是一个有待重点研究的问题。本论文利用分类器进行识别身份证号码,将从两部分介绍身份证号码分类识别方法,第一部分是解释分类器(距离分类器、最小类间距分类器与最大后验概率分类器)的分类公式。第二部分是对图像首先进行主成份分析(PCA)方法进行特征降维,再运用分类器识别身份证号码图像。通过实验结果证明,身份证号码图像经过PCA特征降维后,分类器可以获得很好的分类效果。

Accurate identification of identity card number is a problem needing to be focused on,and there are many applications based on it. At present,there are several ways to identify an identity card number,and in this paper,we used the classifier. Firstly,we introduced the classification formula of the classifiers（ distance classifier,minimum intra-class distance classifier,and maximum a posteriori probability classifier）. Secondly,we reduced the dimensionality of the identity card image by using principal component analysis（ PCA） method,then we used the classifier to identify the identity card number. The experimental results proved that the classifier could obtain better classification effect after PCA dimensionality reduction.