首页 | 本学科首页   官方微博 | 高级检索  
     

基于Adaboost的手写体数字识别
引用本文:赵万鹏,古乐野.基于Adaboost的手写体数字识别[J].计算机应用,2005,25(10):2413-2414.
作者姓名:赵万鹏  古乐野
作者单位:中国科学院,成都计算机应用研究所,四川,成都,610041
摘    要:提出了一种新的基于集成学习算法Adaboost的手写体数字识别系统。Adaboost方法可以在仅比随机预测略好的弱分类器基础上构建高精度的强分类器。实验证明,基于Adaboost的手写体数字识别系统具有较高的识别率和泛化能力,已经应用在OCR识别软件中。

关 键 词:Adaboost  手写体数字识别  弱分类器
文章编号:1001-9081(2005)10-2413-02
收稿时间:2005-04-19
修稿时间:2005-04-192005-07-06

Handwritten digit recognition based on Adaboost
ZHAO Wan-peng,GU Le-ye.Handwritten digit recognition based on Adaboost[J].journal of Computer Applications,2005,25(10):2413-2414.
Authors:ZHAO Wan-peng  GU Le-ye
Affiliation:Chengdu Institute of Computer Application, Chinese Academy of Science, Chengdu Sichuan 610041, China
Abstract:A handwritten digit recognition system based on Adaboost algorithm was introduced in this paper. Adaboost could construct a highly accurate classifier by combining many weak classifiers that just had slightly better accurate than random prediction. Experiment proved that the handwritten system based on Adaboost have low error rate and good ability of generalization. And it was integrated in a oct software.
Keywords:Adaboost  handwritten digit recognition  weak classifier
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《计算机应用》浏览原始摘要信息
点击此处可从《计算机应用》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