| 利用动态核的形态联想记忆网络的研究 |
| |
| 引用本文： | 吴锡生,黄少芳,王士同.利用动态核的形态联想记忆网络的研究[J].计算机工程与设计,2005,26(11):2907-2909. |
| |
| 作者姓名： | 吴锡生 黄少芳 王士同 |
| |
| 作者单位： | 1. 南京理工大学,计算机系,江苏,南京,210094;江南大学,信息工程学院,江苏,无锡,214122
2. 江南大学,信息工程学院,江苏,无锡,214122 |
| |
| 基金项目： | 江苏省自然科学基金项目（BK2003017）. |
| |
| 摘 要： | 在文献1]的基础上，提出了一个基于动态核的形态联想记忆网络方法，特点是同一幅图像，如果其所含的噪声情况不同，则其核也将不同，从而较好地解决了图像含有随机噪声时的联想记忆问题。实验证明，此方法具有良好的性能，双向联想记忆的准确率优于文献1]中介绍的方法。
|
| 关 键 词： | 联想记忆 形态学 形态学神经网络 核动态核 |
| 文章编号： | 1000-7024（2005）11-2907-03 |
| 收稿时间： | 2004-09-26 |
| 修稿时间： | 2004-09-26 |
Results on morphological associative memories based on dynamic kernel |
| |
| WU Xi-sheng,HUANG Shao-fang,WANG Shi-tong.Results on morphological associative memories based on dynamic kernel[J].Computer Engineering and Design,2005,26(11):2907-2909. |
| |
| Authors: | WU Xi-sheng HUANG Shao-fang WANG Shi-tong |
| |
| Affiliation: | 1 Department of Computer, Nanjing University of Science and Technology, Nanjing 210094, China; 2. College of Information Technology, Southern Yangtze University, Wuxi 214122, China |
| |
| Abstract: | The new morphological associative memory based on dynamic kernel is presented. The property of the new method is that the kernel of the same image varies with different noise on it. In the method, the question of associative memories for images with random noise is solved well. The accuracy rate is superior to the method, which is verified by the experiments. |
| |
| Keywords: | associative memories morphology morphological neural networks kernel dynamic kernel |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
