大面积形体三维测量数据拼接技术的研究  

Study on 3D data connection of large part in 3D noncontact measurement

作　　者：王宫[1] 钟约先[1] 袁朝龙[1] 马扬飚[1] 李仁举[1] 

机构地区：[1]清华大学机械工程系先进成形制造教育部重点实验室,北京100084

出　　处：《机械设计与制造》2007年第9期90-92,共3页Machinery Design & Manufacture

摘　　要：大面积形体三维无接触测量,一般采用多视场测量和测量数据拼接的方法,存在积累误差大、测量精度低的问题。采用框架拼接法,通过编码标志点的匹配,求得标志点的空间三维坐标,完成物体空间框架的重构;然后通过局部三维扫描点云与空间框架之间的特征点的提取,得到一组匹配点对,运用SVD矩阵分解算法求出旋转矩阵R和平移向量T,实现局部三维扫描数据的精确拼接。实例表明,该方法有效地提高了大面积形体三维测量的精度。The method that it measures from different angles and connects together is commonly adopted in large part＇s 3D noncontact measurement, which has the problems of great accumulated error and low precision. This paper produces a method which connects part＇3D scanning cloud data to a frame. In this method, marked points＇3D coordinates are calculated, which form the frame. Then the feature points between the patr＇3D scanning cloud data and the frame are identified and matched. And the rotation matrix R and the translation vector T are found through the singular value decomposition （SVD） algorithm. Thus, the connection solution is achieved. A real example indicates that the method improves the precision of the large part＇s 3D measurement.

关 键 词：三维测量 测量数据拼接 标志点 匹配 

分 类 号：TH16[机械工程—机械制造及自动化]