摘要
点云数据三角化处理是逆向工程及快速原型领域中不可缺少的环节。针对传统点云数据三角化算法效率低及鲁棒性差的特点,结合面结构光三角法测量原理,提出融合图像域信息的点云数据三角化方法;根据点云数据图像域信息所具有的空间拓扑关系完成图像域内的三角化连接;然后将其映射到三维空间域中完成点云数据的三角化。由于图像域具有拓扑确定性,不需要将点云数据映射到二维平面再进行平面空间的拓扑关系确定及三角化处理,从而使算法的鲁棒性及处理效率得到提升,最后以具体的实例验证算法有效性。算法已用到具体的项目中实现了测量过程中点云数据的快速三角化。
It is necessary to triangulate the point cloud in reverse engineering and rapid prototyping. The efficiencies of traditional triangulation methods are low and their robustness is bad. A method of triangulation using the image field information is proposed in the procedure of structural light measurement. The method uses relationship of pixel in image field to describe the relationship of 3D point cloud. Due to the fact thtat pixels in image field have definite topology, it is not necessary to map 3D point cloud to 2D domain for building their topology relationship. Therefore, the efficiency and robustness of the triangulation procedure is improved. Finally, one case is used to prove the validity of the method. The method has already been applied to engineering practice.
出处
《机械科学与技术》
CSCD
北大核心
2008年第4期421-424,共4页
Mechanical Science and Technology for Aerospace Engineering
基金
宁波市自然科学基金项目(2007A610011)资助
关键词
点云数据
三角化
图像域
鲁棒性
结构光测量
point cloud
triangulation
image field
robustness
structural light measurement
