|
|||||
|
|
| 反求工程中基于点云数据集的CAD建模研究 | |
| 引用本文： | 鄢腊梅,孙晓,周锋. 反求工程中基于点云数据集的CAD建模研究[J]. 机械设计与研究, 2005, 21(6): 72-74 |
| 作者姓名： | 鄢腊梅 孙晓 周锋 |
| 作者单位： | 湖南工业大学,机械学院,株洲,412008;清华大学,自动化系,北京,100084 |
| 基金项目： | 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，同时能满足反求工程的实时需求. |
| 关 键 词： | 光顺 神经网络 反求工程 点云 |
| 文章编号： | 1006-2343(2005)06-072-03 |
| 收稿时间： | 2005-03-01 |
| 修稿时间： | 2005-03-01 |
A New Approach to CAD Model Reconstruction From Cloud Datas Using Neural Networks |
|
| YAN La-mei,SUN Xiao,ZHOU Feng. A New Approach to CAD Model Reconstruction From Cloud Datas Using Neural Networks[J]. Machine Design and Research, 2005, 21(6): 72-74 | |
| Authors: | YAN La-mei SUN Xiao ZHOU Feng |
| Affiliation: | 1. School of mechanical Engineering, Hunan University of polytechnic,Zhu Zho, 412008,China; 2. Department of Automation,Tsinghua University,Bejing 10084,China |
| Abstract: | 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 extracted 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. |
| Keywords: | smooth,neural networks,reverse engineering cloud datas |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |