摘要
在这篇文章中，我们将介绍一种新型电子电路图纸自动读入系统的构成与实现。这是一个已被我们研制成功的实验系统。下面将扼要介绍其中的模式分离与图象识别技术。对于前者，我们着重介绍基于Ｆｒｅｅｍａｎ链码边缘跟踪技术的字符分离方法；而后者则包括两方面的内容：实现字符识别的人工神经网络方法与实现符号识别的决策树方法，最后给出了经识别后的输出图形。
In this paper we presented a new structure of automatical read - in system for the electronic circuit drawings. It was a prototypy which has been developed by us recently' This system was different from others.We described the methods of the pattern separating and image recognition principally. For the former,we have discussed the character separating based on the edge -trace technique of Freeman chain -code, for the latter, we have presented a character recognition method based on the artificial neural networks and a symbol recognition method of electronic circuit based on the 'decision tree'' Finally, the output graphy after recognition was given.
出处
《微电子学与计算机》
CSCD
北大核心
1996年第3期22-26,共5页
Microelectronics & Computer
基金
国家高技术863计划项目
自然科学基金
关键词
电子电路图
模式分离
图象识别
决策树
Binary images
Pattern separating
Image recognition
Artificial neural networks
Decision tree
