首页 | 官方网站   微博 | 高级检索  
     

手写汉字识别的非线性动态部件模板
引用本文:石大明,刘家锋,唐降龙,舒文豪.手写汉字识别的非线性动态部件模板[J].自动化学报,2004,30(3):390-399.
作者姓名:石大明  刘家锋  唐降龙  舒文豪
作者单位:1.南洋理工大学计算机工程学院,新加坡;
摘    要:对以前提出的非线性动态手写模板加以改进并用于手写汉字的部件识别.在训练阶 段,核-主元分析用来捕捉非线性的手写变化.于是,只需改变少量的形状参数就可获得动态变 形的模板.在识别阶段,遗传算法取代了原始的动态通道算法去寻找最优的形状参数.我们对覆 盖2154个汉字类别的200个部件进行了实验,对不用人书写的430,800个测试样本的部件识 别率达97.4%.与现有的代表性部件方法比较也显示本文的方法效果最好.

关 键 词:手写汉字识别    动态手写模板    核-主元分析    遗传算法
收稿时间:2003-2-21
修稿时间:2003年2月21日

Nonlinear Active Radical Modeling for Handwritten Chinese Character Recognition
SHI Da-Ming,LIU Jia-Feng,TANG Xiang-Long,SHU Wen-Hao.Nonlinear Active Radical Modeling for Handwritten Chinese Character Recognition[J].Acta Automatica Sinica,2004,30(3):390-399.
Authors:SHI Da-Ming  LIU Jia-Feng  TANG Xiang-Long  SHU Wen-Hao
Affiliation:1.Dschool of Computer Engineering,Nanyang Technological University,Singapore;Department of Computer Science,Harbin Institute of Technology,Harbin
Abstract:This paper improves the authors' previously proposed nonlinear active hand-writing models and applies them into radical extraction for handwritten Chinese charac-ter recognition. In the training phase, kernel principal component analysis is used to capture nonlinear handwriting variations. Then deformable models can be generated by varying a small number of shape parameters. In the recognition phase, genetic algo-rithms, rather than dynamic tunneling algorithm in the original version, are employed to search for the optimal shape parameters. Experiments are conducted on 200 radicals covering 2154 character categories. The correct matching rate is 97. 4% on 430,800 loosely-constrained characters. Comparison with existing representative radical approa-ches shows that our method achieves superior performance.
Keywords:Handwritten Chinese character recognition  active handwriting models  kernel principal component analysis  genetic algorithms
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《自动化学报》浏览原始摘要信息
点击此处可从《自动化学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号