| 基于稀疏RAM的N-tuple神经网络的足球机器人识别法 |
| |
| 引用本文： | 彭金华.基于稀疏RAM的N-tuple神经网络的足球机器人识别法[J].江西电力职业技术学院学报,2006,19(2):41-42,62. |
| |
| 作者姓名： | 彭金华 |
| |
| 作者单位： | 东南大学无线电工程系,江苏,南京,210096 |
| |
| 摘 要： | 提出了一种采用稀疏RAM的N-tuple神经网络(SN-tuple)的方法,用在足球机器人比赛时,根据图像快速识别队员。通过大量实验证明,该方法有效地解决了传统BP神经网络收敛速度慢、易陷入局部最小值的问题,对机器人队员和足球位置的识别稳定可靠。
|
| 关 键 词： | 足球机器人 N-tuple神经网络 机器视觉 |
| 文章编号： | 1673-0097（2006）02-0041-（02） |
| 收稿时间： | 03 9 2006 12:00AM |
| 修稿时间： | 2006-03-09 |
Football Robot Recognition Method Based On N-tuple Neural Network with the Rarefaction RAM |
| |
| PENG Jin-hua.Football Robot Recognition Method Based On N-tuple Neural Network with the Rarefaction RAM[J].Journal of Jiangxi Vocational and Technical College of Electricity,2006,19(2):41-42,62. |
| |
| Authors: | PENG Jin-hua |
| |
| Institution: | Department of Radio Engtneering, Southeast University, Nanjing 210096, China |
| |
| Abstract: | In this paper,a new approach,named N-tuple neural network with the rarefaction RAM,is presented.It is used in the football robot match,can swith recognize the members of the two teams.Experimental results demonstrate,in this way,it could overcome the shortage of using BP network,such as refrain speed slow and easy fall into minimum,the performance is increased and more stable and reliable. |
| |
| Keywords: | football robot N-tuple neural network with the rarefaction RAM machine vision |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
