一种基于支持向量机的手写汉字识别方法

本文提出了一种新的基于支持向量机手写汉字识别方法.支持向量机作为一种新的机器学习方法,由于其建立在结构风险最小化准则之上,而不是仅仅使经验风险达到最小,从而使得支持向量分类器具有较好的推广能力.本文首先讨论了支持向量机的基本原理,然后,针对支持向量机识别大类别手写汉字所遇到的特殊问题,文章进行了分析和阐述,并在此基础上,提出了基于最小距离分类器预分类的两级分类策略.最后,针对GB2312-80的1034个汉字类别的120套手写样本,进行了实验仿真.实验结果表明,本文方法的汉字识别率较距离分类器有较大提高,其中多项式核函数的支持向量分类器,识别率平均提高3.38％,表明了本文方法的有效性.

A New SVM-Based Handwritten Chinese Character Recognition Method

A new recognition method of handwritten Chinese characters by support vector machine is presented.Support vector machines(SVM) operate on the principle of structure risk minimization which not only keeps the empirical risk minimal but also controls VC confidence of discriminant functions,hence a better generalization ability is guaranteed.In this paper,the problems to be solved while applying SVM in Chinese character recognition are addressed at first,and then a two stage of recognition scheme is suggested.Finally,experimental results on 1034 categories of Chinese character from 120 sets of samples are given.For the Polynomial kernel SVM,a 3 38% average increment of recognition rate is obtained showing the efficiency of the proposed approach.