电子电路图的自动输入、理解及与ORCAD的链接

本文介绍一个最近由我们研制成功的电子电路图自动输入系统的构成与特点。利用本文提出的内、外边缘快速判决法及在“非环形”符号各端点加“黑点”的辅助措施提高了模式分离的可靠性。为了最大限度地利用原始信息，我们采用直接从光栅图中抽取特征的办法去识别字符与符号。在手写符号的识别中，我们提出了四边带投影的特征抽取方法并把它了引入到“决策树”的判决过程中，使识别率与识别速度都得到提高。当系统对整图的拓扑关系进行理解之后，还实现了与当前国际流行的ＣＡＤ软件ＯＲＣＡＤ的链接、使系统能以自动输入的方式支持自动布线与电路模拟。

Automatic Input, Understanding and Linking with ORCAD for Electronic Circuit Diagrams

This paper describes the structure and features of an automatic input system for the electronic circuit diagrams. By means of the rapid detecting method to tell the external contours from the internal ones and with the aid of adding" black points" on the terminals of each non-circular symbols, the reliability of pattern separation is increased. In order to make the best use of the primtive information, the feature extraction of both element symbols and characters is based on the raster data. For the handwritten symbols, recognition we proPOse an extracting feature method projecting four sidebands and bring it in the decision procedure of the decision tree. As a result,both the recognition rate and speed have obviously increased. When the system understands the topologic relationship of the whole graph, it can link with ORCAD which is a current international CAD software. Thus, this system can support the line routing of printed circuit board and circuit simulation in the manner of automatic input.