基于人工神经网络识别的特征自组织技术

THE METHOD FOR FEATURE SELF-ORGANIZATION BASED ON FEATURE RECOGNITION WITH NEURAL NETWORK

摘要: 特征识别是CAD中一项重要内容,也是一类智能化程度较高的图形模式识别问题.神经网络的智能性和并行性使它成为特征识别的最新方法.在特征的扩展型面邻接图表达方法基础上,运用神经网络识别技术,提出并实现了特征的自组织技术,从而为特征造型和特征识别提供了一种新的有效方法.

Abstract: Feature recognition is an important topic in CAD，and is an intelligent and different problem in pattern recognition．In this paper，on the basis of feature representation with Extensive Face Adjacent Graph（EFAG），the method of feature recognition with neural network is presented and feature self-organization is implemented．It provides a new method for feature modelling and feature recognition．