首页 | 本学科首页   官方微博 | 高级检索  
     

一种新的神经网络空洞滤波器的实现
引用本文:汪海明,郭仕德,余道衡. 一种新的神经网络空洞滤波器的实现[J]. 电路与系统学报, 2003, 8(3): 6-9
作者姓名:汪海明  郭仕德  余道衡
作者单位:1. 北京大学电子学系视觉与听觉信息处理国家重点实验室,北京,100871
2. 北京大学遥感与地理信息系统研究所,北京,100871
基金项目:国家863基金资助项目(2002AA716081)
摘    要:本文深入研究了一种新的二维细胞自动机(CA),并据此提出了一种新的神经网络空洞滤波器实现方法,给出了系统实现的原理电路,并称其为细胞神经网络空洞滤波器(CNN-HF)。仿真结果表明,此系统能够有效的实现二值字符的空洞填充,且具有运行速度快,硬件实现容易等特点,期望能够在二维CA研究与应用设计方面有所启发和突破。

关 键 词:细胞自动机(CA) 细胞神经网络(CNN) CNN空洞滤波器 (CNN-HF)
文章编号:1007-0249(2003)03-0006-04
修稿时间:2002-09-02

Realization of New Hole Filter Based on CNN
WANG Hai-ming,GUO Shi-de,YU Dao-heng. Realization of New Hole Filter Based on CNN[J]. Journal of Circuits and Systems, 2003, 8(3): 6-9
Authors:WANG Hai-ming  GUO Shi-de  YU Dao-heng
Affiliation:WANG Hai-ming1,GUO Shi-de2,YU Dao-heng1
Abstract:Based on the discussion of two-dimensional Cellular Automata (CA), a new kind of hole filter based on Cellular neural Networks, named as CNN-hole filter (CNN-HF) is proposed. Related circuit is accordingly designed to implement its function. Simulation results show that using the proposed algorithm, the hole in a figure can be filled effectively. Furthermore, it has many advantages, such as, high-speed processing, simple structure and easy integration and so on. It is expected to have profound influence on the technology of two-dimensional Cellular Automata.
Keywords:Cellular Automata (CA)  Cellular neural network (CNN)  Hole-filter based on CNN (CNN-HF)  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号