English

应用科学学报 ›› 1999, Vol. 17 ›› Issue (3): 296-301.

• 论文 • 上一篇    下一篇

混沌神经网络用于灰度图象的联想记忆

李克, 杨绿溪, 何振亚   

  1. 东南大学
  • 收稿日期:1997-12-18 修回日期:1998-11-12 出版日期:1999-09-30 发布日期:1999-09-30
  • 基金资助:
    国家自然科学基金、教育部博士点青年基金和江苏省自然科学基金资助项目

Associative Memory of Gray-Scale Images with Chaotic Neural Networks

LI KE, YANG LÜXI, HE ZHENYA   

  1. Department of Radio Engineering, Southeast University, Nanjing 210018
  • Received:1997-12-18 Revised:1998-11-12 Online:1999-09-30 Published:1999-09-30

PDF

68

被引次数

35

摘要： 针对一般联想记忆系统只能处理二值模式的缺点提出一种多值模式的相关学习方法,并在改进的GCM模型(S-GCM)基础上实现了对多灰度级图象的记忆与识别,最后对识别的效果和影响因素作了分析.

关键词: 联想记忆, 全局耦合映射, 混沌神经网络, 模式

Abstract: In this paper, we propose a novel correlative learning rule for multi-value patterns, considering that conventional associative memory system processes only binary ones. Successful memory and recognition are achieved on the bases of S -GCM model with the proposed method, and analyses on the recognition results and factors are given.

Key words: chaotic neural networks, associative memory, globally coupling map, pattern