|
|||||
|
|
| 基于忆阻脉冲耦合神经网络的图像去噪 | |
| 引用本文： | 高宏宇,黄文丽,董宏丽,李佳慧.基于忆阻脉冲耦合神经网络的图像去噪[J].吉林大学学报(信息科学版),2020,38(1):49-54. |
| 作者姓名： | 高宏宇 黄文丽 董宏丽 李佳慧 |
| 作者单位： | 东北石油大学a． 电气信息工程学院; b． 黑龙江省网络化与智能控制重点实验室，黑龙江大庆163318 |
| 基金项目： | 国家自然科学基金资助项目( 61873058) ; 中国博士后基金资助项目( 2017M621242) ; 黑龙江省自然科学基金资助项目 ( F2018005) |
| 摘 要： | 为解决传统脉冲耦合神经网络的参数不固定问题，在图像处理中应用忆阻元件突出的记忆属性，提出了
应用两个忆阻元器件反向并联模拟脉冲神经网络中的神经元间的连接强度，构建新型忆阻脉冲神经网络，实现
神经元间连接强度动态可变化，再将该新型网络用于图像去噪问题。通过Matlab 仿真实验，验证了改进后的
新型网络在图像去噪方面的良好性能，并通过峰值信噪比和图像相似度指标证明了该方法用于图像去噪具有
较好的效果。 |
| 关 键 词： | 忆阻器 脉冲耦合神经网络 图像去噪 |
| 收稿时间： | 2019-07-31 |
Image Denoising Based on Memristive Pulse Coupled Neural Network |
|
| GAO Hongyu,HUANG Wenli,DONG Hongli,LI Jiahui.Image Denoising Based on Memristive Pulse Coupled Neural Network[J].Journal of Jilin University:Information Sci Ed,2020,38(1):49-54. | |
| Authors: | GAO Hongyu HUANG Wenli DONG Hongli LI Jiahui |
| Institution: | a． School of Electrical Engineering and Information; b． Heilongjiang Provincial Key Laboratory of Networking and Intelligent Control，Northeast Petroleum University，Daqing 163318，China |
| Abstract: | In order to solve the parameter immobilization of the traditional pulse-coupled neural network，the memory properties of the memristive components are applied in the image processing，and two memristive components in the anti-parallel are used in the connection strength analog between the neurons of pulse neural network． A novel memristive pulse neural network is constructed to realize the dynamic change of the connection strength between neurons，and the new network is used for image denoising． The Matlab simulation experiment is carried out to verify the good performance of the improved new network in image denoising． The peak signal-tonoise ratio and image similarity index prove that the method has good effect on image denoising． |
| Keywords: | memristor pulse coupled neural networks image denoising |
| 点击此处可从《吉林大学学报(信息科学版)》浏览原始摘要信息 | |
| 点击此处可从《吉林大学学报(信息科学版)》下载免费的PDF全文 | |