摘要： 提出了密集测量数据云中规则表面自动识别的通用算法,该算法以相邻测点组成的三角片法向变化来搜索规则表面及其边界,可应用于反求工程及各种机械仿形建模加工软件中。对于精度不高的测量数据云,可利用二维密度算法提取规则表面。该算法已成功应用于某商业用激光机器视觉三维检测系统中。-

关键词: 规则表面自动识别, 三维测量数据云, 二维密度, 反求工程

Abstract: Presents a practical general algorithm to automatically sort 3D datacloud into various simple feature primitives such as planes and cylinders, by using the triangle normal direction composed by the neighbored points. Also presents a new concept DataCloud 2D Density to sort planes and cylinders.

中图分类号: 

• TP391.41