摘要
提出了一个应用于广泛信息源的立体模型重构及应用系统,它基于规则的专家系统,主要由输入资料处理、图形事实获取、知识库、造型处理和特征处理等五部分组成。根据输入资料源的不同,采用相应技术进行识别和理解,完成图形事实的获取。基于规则的专家系统模拟人工读图的理解过程,由读图知识,经验规则推理出组合体的所有基本形体及组合形式。重构主体模型中的特征识别和提取技术为CAD/CAM的处理提供了全面的信息。
The paper presents a rule-based system of 3D-model reconstruction and application. It can use considerably broad information resources. The system consists of five parts: input information , processing extraction of graphics, facts knowledge base, reconstruction of geometric entities and feature modeling. The gaining graphics are performed by different identification - and comprehension techniques based on different kinds of information sources. The role-based system can imitate human's logical thought process. All kinds of possible single figures and their combination figures are reasoned by drawings' knowledge and experience principles. In the 3D-model reconstruction, the techniques of feature identification and extraction provide comprehensive useful information for CAD/CAM.
出处
《工程图学学报》
CSCD
2003年第2期44-51,共8页
Journal of Engineering Graphics
