| 具有抗随机噪声能力的联想记忆网络 |
| |
| 引用本文： | 冉天保,吴锡生,王士同. 具有抗随机噪声能力的联想记忆网络[J]. 计算机工程与设计, 2007, 28(11): 2662-2664 |
| |
| 作者姓名： | 冉天保 吴锡生 王士同 |
| |
| 作者单位： | 江南大学,信息工程学院,江苏,无锡,214122;无锡商业职业技术学院,信息系,江苏,无锡,214153;江南大学,信息工程学院,江苏,无锡,214122 |
| |
| 摘 要： | 形态联想记忆网络具有良好的联想记忆功能和较强的抗膨胀或腐蚀噪声能力,但抗混合噪声的能力很弱.而在实际中,随机噪声往往是混合型的,既有膨胀又有腐蚀噪声,将尺度空间和形态联想记忆网络相结合,得到了一种新的联想记忆网络,它提高了形态自联想记忆网络的抗随机噪声能力.通过仿真实验验证了该方法具有良好的性能.
|
| 关 键 词： | 形态学 尺度空间 自联想记忆 膨胀存储矩阵 腐蚀存储矩阵 |
| 文章编号： | 1000-7024（2007）11-2662-03 |
| 修稿时间： | 2006-05-18 |
Morphological associative memories of resisting random noise |
| |
| RAN Tian-bao,WU Xi-sheng,WANG Shi-tong. Morphological associative memories of resisting random noise[J]. Computer Engineering and Design, 2007, 28(11): 2662-2664 |
| |
| Authors: | RAN Tian-bao WU Xi-sheng WANG Shi-tong |
| |
| Affiliation: | 1. School of Information Engineering, Southern Yangtze University, Wuxi 214122, China; 2. Department of Information Engineering, Wuxi Institute of Commerce, Wuxi 214153, China |
| |
| Abstract: | Morphological associative memories(MAM)has strong associative memorizing ability and fairly strong capability against dilative or erosive noise,but its ability against general noise is poor.But in practice,random noise is always composed both of dilative and erosive noise.It is proposed that scale space and MAM is combined to get a new version of MAM with improved noise resisting ability.The results of emulation experiment testify it good performance. |
| |
| Keywords: | morphology scale space auto-associative memories dilative memory matrix erosive memory matrix |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
