一种点云分割金属铸锭的定位方法

A positioning method for point cloud segmentation of metal casting ingots

针对AGV叉车搬运金属铸锭过程中的精准定位问题,提出了一种金属铸锭表面三维点云分割算法,主要包括预处理和点云分割两个步骤。预处理步骤对初始点云进行压缩和精简,点云分割步骤采用点云重分配方式改进RANSAC算法,解决了分割过程中的错分割和欠分割问题。实验结果表明,算法对目标金属铸锭三维点云分割具有较好的分割效果且用时较短。

Aiming at the problem of precise positioning in the process of AGV forklifts transporting metal ingots,a three-dimensional point cloud segmentation algorithm on the surface of metal ingots is proposed,which mainly includes two steps:preprocessing and point cloud segmentation.Through preprocessing compress and streamlines of the initial point cloud,it uses point cloud redistribution to improve the RANSAC algorithm,which solves the problems of wrong segmentation and under-segmentation during the segmentation process.Experimental results show that the algorithm has a good segmentation effect and takes a short time for the three-dimensional point cloud segmentation of target metal ingots.