| 细胞神经网络与字符特征提取技术研究 |
| |
| 引用本文： | 费文东,孟相如.细胞神经网络与字符特征提取技术研究[J].空军工程大学学报,2000,1(3):51-54. |
| |
| 作者姓名： | 费文东 孟相如 |
| |
| 作者单位： | 空军工程大学电讯工程学院!陕西西安710077 |
| |
| 摘 要： | 首先对细胞神经网络 ( CNN)的基本模型作了简要论述 ,在对连续时间 CNN离散化的基础上 ,将连通单元检测与阴影检测相结合 ,提出了一种基于 CNN文字识别的特征提取技术 ,并进行了计算机的仿真试验 ,得到了良好的结果。
|
| 关 键 词： | 细胞神经网络 连通单元检测 阴影检测 特征提取 |
| 修稿时间： | 1999-12-17 |
Studies on the Cellular Neural Networks and Character Feature Extraction Techniques |
| |
| FEI Wen-dong,MENG Xiang-ru.Studies on the Cellular Neural Networks and Character Feature Extraction Techniques[J].Journal of Air Force Engineering University(Natural Science Edition),2000,1(3):51-54. |
| |
| Authors: | FEI Wen-dong MENG Xiang-ru |
| |
| Institution: | The TelecommumcatlOn Engineermg Inst1tute ,Xi''an 710077 ,China |
| |
| Abstract: | The basic model of the cellular Neural Network(CNN) is briefly described in this paper.Based on the discretetransformation of continuous time CNN,feature extraction techniques of character recognition using CNN are proposed by the combination of connected component detector and shadow detector.Good results of computer simulation experiment are also given. |
| |
| Keywords: | cellular neural network connected component detector shadow detector feature extraction |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《空军工程大学学报》浏览原始摘要信息 |
| 点击此处可从《空军工程大学学报》下载免费的PDF全文 |
