摘要
本文研究了一种适用于旋转盘3D扫描系统的三维点云快速准确配准的算法问题,利用飞行时间技术提高原始三维点云数据采集精度,减少源头数据误差引入,通过辅助的参照物多图扫描获取旋转轴信息及其转换矩阵,为目标物体的三维点云粗配准提供较好的初始信息,继而实现高精度、快速的目标物体的点云精配准,为高精度的增材制造过程提供算法支撑。
This paper studies an improved iterative closest point (ICP) algorithm of fast and accurate registration of 3D point cloud for the rotating disk 3D scanning system. The author first adopts the time-of-flight technology to improve the acquisition accuracy of original 3D point cloud data sampling, reducing the introduction of source data errors. Then, she uses an auxiliary reference object to scan multiple images, and obtain the rotation axis information and its transformation matrix, in order to provide better initial information for the rough registration of the 3D point cloud of the target object. Finally, the high-precision and fast point cloud registration of the target object is achievable, which makes a high-precision additive manufacturing process available.
出处
《计算机科学与应用》
2022年第4期1023-1030,共8页
Computer Science and Application
