| 引用本文： |
-
谢晓兰,韩可轶,冯嘉礼.基于JAVA的鼠标书写的非常用字符识别[J].广西科学院学报,2006,22(1):60-63. [点击复制]
- XIE Xiao-lan,HAN Ke-yi,FENG Jia-li.Recognition of Non-common Characters by Mousewriting Based on JAVA[J].Journal of Guangxi Academy of Sciences,2006,22(1):60-63. [点击复制]
|
|
| 摘要: |
| 采用JAVA语言,通过合理设置网络的变量,改进误差逆传播校正方法(BP算法),缩短了BP算法的学习时间,提高算法的运算速度,实现了鼠标书写的非常用字符识别.试验程序验证表明,识别程序能快速准确识别用鼠标书写的字符. |
| 关键词: 字符识别 非常用字符 鼠标 神经网络 BP算法 JAVA |
| DOI： |
| 投稿时间：2005-05-08 |
| 基金项目:广西教育厅科研项目(桂教科研[2003]22号)资助 |
|
| Recognition of Non-common Characters by Mousewriting Based on JAVA |
|
XIE Xiao-lan1, HAN Ke-yi1, FENG Jia-li2
|
| (1.Department of Electronics and Computer Science, Guilin University of Technology, Guilin, Guangxi, 541004, China;2.Infomation Engineering College, Shanghai Maritime University, Shanghai, 200035, China) |
| Abstract: |
| Java language and BP algorithm are used to develop a method to identify the writing characters by using a mouse.The BP learning time is getting shorter by improvement of BP algorithm and proper setting of network parameters.In the examination,by using the deducing of theory and the demonstrating of programming,the error and the rate of identification are acceptable for usual inputting. |
| Key words: character recognition non-common characters mouse neural network BP algorithm JAVA language |
