反求工程中基于点云数据集的CAD建模研究  

A New Approach to CAD Model Reconstruction From Cloud Datas Using Neural Networks

作　　者：鄢腊梅[1] 孙晓[1] 周锋[2] 

机构地区：[1]湖南工业大学机械学院,株洲412008 [2]清华大学自动化系,北京100084

出　　处：《机械设计与研究》2005年第6期72-74,共3页Machine Design And Research

基　　金：国家自然科学基金资助项目(60173046)

摘　　要：提出了一种新的基于神经网络的点云数据重构CAD光顺造型的新算法。首先对点云数据平滑处理;然后进行特征线提取,并以特征线为基础对曲面进行分割。该方法能直接从神经网络的权值矩阵得到曲线的控制顶点/曲面的控制网格,通过神经网络的权值约束实现曲线段/曲面片之间的光滑拼接。同时对恢复的隐式表面的初始逼近网格自适应性优化。实验效果表明,该方法能够得到精确的逼近结果。The paper presents a new approach to CAD model reconstruction from cloud datas using neural networks（NN）. It includes three steps：first, the cloud data are preprocessed for smoothing;second, feature lines are ex tracted and the cloud data are segmented; at last, implicit surface patches are created over rectangular mesh and trimmed to form an entire surface using RBFNN（Radical basis function neural networks）. And then the triangles in the coarse mesh are recursively subdivided by employing self-adaptive method put forward. Experimental results show that this approach can get more accurate result of approximation. The rendering and control of CAD model can be treated in real-time reverse engineering.

关 键 词：光顺 神经网络 反求工程 点云 

分 类 号：TH131.41[机械工程—机械制造及自动化]