| 点状符号识别的旋转、缩放不变性研究 |
| |
| 引用本文： | 宋伟东,何孝莹.点状符号识别的旋转、缩放不变性研究[J].测绘与空间地理信息,2007,30(4):19-20,24. |
| |
| 作者姓名： | 宋伟东 何孝莹 |
| |
| 作者单位： | 辽宁工程技术大学,辽宁,阜新,123000 |
| |
| 摘 要： | 针对点状符号神经网络识别中的旋转、缩放不变性问题,利用不变性特征输入和构建旋转、缩放不变性网络两种方法加以实现。前者的关键是提取符号的不变性特征作为网络的输入,后者的重点是构造复合神经网络实现不变性识别分类。
|
| 关 键 词： | 特征 网络模型 识别 |
| 文章编号： | 1672-5867（2007）04-0019-02 |
| 收稿时间： | 2006-10-04 |
| 修稿时间： | 2006-10-04 |
The Research of Rotating and Zooming Fixity in Recognition of Mark Map Symbols |
| |
| SONG Wei-dong,HE Xiao-ying.The Research of Rotating and Zooming Fixity in Recognition of Mark Map Symbols[J].Geomatics & Spatial Information Technology,2007,30(4):19-20,24. |
| |
| Authors: | SONG Wei-dong HE Xiao-ying |
| |
| Institution: | Liaoning Technical University, Fuxin 123000, China |
| |
| Abstract: | Two methods are realized to reach the fixity results when rotating and zooming symbols in symbols recognition.The former is to get the characters which do not change when being rotated or zoomed,then these characters are put into network to recognize symbols.The later is to build a special compound net which can recognize symbols without focusing on the changing of mark. |
| |
| Keywords: | characters network model recognize |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
