节点文献

面向RPM的点云数据直接分层处理技术及其软件实现

【作者】 陈伟卿

【导师】 吴宏基; 孙玉文;

【作者基本信息】 大连理工大学 , 机械制造及其自动化, 2002, 硕士

【摘要】 本文研究了反求工程与快速原型制造集成中的关键问题,即如何从点云数据生成层片轮廓数据。目前主要有两种方法:一是建立点云数据的CAD模型,并对该模型三角剖分为STL文件,从STL文件中生成切片数据;二是直接将点云数据离散成STL文件,以生成切片。但是这两大类方法都存在消耗时间长、人工交互多等缺陷。 为了克服上述方法的不足,提出了一种新的方法,即点云数据的直接分层技术,主要包括数据预处理、点云数据直接分层方法、轮廓追踪与拟合三方面内容。 本文首先阐述了点云数据的预处理方法,包括选择分层位向、建立拓扑关系、法向补偿及平滑处理等操作;其次研究了直接分层的两种方法,即平行平面序列分层法和自由曲线法平面序列分层法,二者的基本原理相同,都是根据最小距离点对的连线与分层面截交求得切片数据;最后介绍了轮廓数据的处理,包括体元轮廓的追踪和轮廓线的提取,以及用双圆弧对轮廓线进行拟合等。

【Abstract】 This paper studies how to generate the slice data from point clouds, which is the critical problem in the integration of Reverse Engineering and Rapid Prototyping Manufacturing. At present, there are two ways available. In the firstpath, a CAD model is created first from point clouds then converted to a STL file. The STL file is then sliced to generate a series of layers for RP fabrication. The second path is to create the STL file directly from point clouds. However, there are drawbacks in these methods, such as long tune consumed, much manual interaction and so on.A new method is brought forward to solve this question. It goes directly from point clouds to a RP slice file. Three problems, the preprocess of point clouds, the method of directly slicing the point clouds, the extraction and fit of contour data, are discussed.This paper first expatiates the preprocess of data, which includes choosing the orientation to slice, topologizing the data, offsetting in the normal direction and smoothing the data. Secondly, it studies two ways of directly slicing, which is to generate the contour data on a series of parallel planes, or normal planes of a freeform curve. These two methods have the same principle, which gets the contour data by crossing the plane and the line between the two points of least distance. Finally, it addresses the process of contour data, such as extracting a cell-contour, sorting the points in one cell-contour to a contour line, and fitting the contour to a smooth one with biarc curves.

  • 【分类号】TP311.52
  • 【被引频次】5
  • 【下载频次】213
节点文献中: 

本文链接的文献网络图示:

本文的引文网络