| 手写体字符识别的多特征多分类器设计 |
| |
| 引用本文： | 沈淑娟,姜建国,曹建春. 手写体字符识别的多特征多分类器设计[J]. 计算机工程与应用, 2004, 40(16): 116-118 |
| |
| 作者姓名： | 沈淑娟 姜建国 曹建春 |
| |
| 作者单位： | 西安电子科技大学计算机学院,西安,710071;西安电子科技大学计算机学院,西安,710071;西安电子科技大学计算机学院,西安,710071 |
| |
| 摘 要： | 特征选取和分类器设计是字符识别系统设计的关键。文章针对手写体汉字和阿拉伯数字混和字符集的识别提出了依据不同的分类要求,分别选取不同的字符特征并采用神经网络多分类器进行识别的设计方法。实验结果表明,该方法用于手写体混合字符集的识别是行之有效的。
|
| 关 键 词： | 特征选取 神经网络 多分类器 手写体字符识别 |
| 文章编号： | 1002-8331-(2004)16-0116-03 |
Design of Multiple Features and Multiple Classifiers for Handwritten Character Recognition |
| |
| Shen Shujuan Jiang Jianguo Cao Jianchun. Design of Multiple Features and Multiple Classifiers for Handwritten Character Recognition[J]. Computer Engineering and Applications, 2004, 40(16): 116-118 |
| |
| Authors: | Shen Shujuan Jiang Jianguo Cao Jianchun |
| |
| Abstract: | The selection of feature sets and the design of classifiers are at the heart of a character recognition system.A novel method for the recognition of handwritten Chinese characters and digits is presented in this paper.Considering diverse classification requirements,different features are selected and an integration of multiple neural networks is uti-lized.The experimental results obtained demonstrate that the proposed approach is promising. |
| |
| Keywords: | feature selection neural network multiple classifiers handwritten character recognition |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
