多视点云数据自动拼合算法  

The automatic merging algorithm of Multi-View point cloud data

作　　者：周波[1] 陈新河[1] 谢雪冬[1] 

机构地区：[1]黑龙江科技学院,哈尔滨150027

出　　处：《自动化与仪器仪表》2007年第1期11-12,32,共3页Automation & Instrumentation

摘　　要：通过在被测物体加贴参考点,提出了以参考点的距离空间不变量为依据来搜索同构子图的方法来对多视中的标记参考点进行对应匹配;通过SVD的方法求解多视间的坐标转换矩阵以来最终达到多视间拼合。运用实例检验表明本文的算法快速、简洁,同时,拼合的鲁棒性也很好。在一定的拼合的精度下,提高了运行的速度和算法的稳定性。Adding reference points on object surface which will be measured belorenana, tins article put forward： First, based on space distance-invariable, searching isomorphic child map to find out matching between reference points those in different views. Then, by SVD method, figure out translation matrix and rotation matrix between different views, And finally put all the point in the same coordinate. The experiment shows that the paper＇ s method is swift, succinctly and has good robustness . So this method has improved the speed and stability of merging algorithm on a certain precision degree.

关 键 词：拼合 对应匹配 数据点云 SVD算法 多视变换 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]