摘要
对手写体汉字识别在特征提取、分类识别及后处理三个阶段主要采用的方法做了简要介绍。SVM具有结构简单,分类稳定可靠,且容错性好等优点。同时和LVQ神经网络分类器识别方法进行了比较,计算机仿真表明,采用SVM用于手写体汉字识别更适合。
This paper surveys the main techniques in three phases：feature extraeting,classification and recognition ,and postproeessing The structure of support vector machine is simple, steady , error-tolerated. The results are also compared with LVQ neural network elassifter . The computer simulating shows that using support vector machine to discern the handwritten Chinese characters is preferable.
出处
《计算机与数字工程》
2006年第1期65-68,共4页
Computer & Digital Engineering
