机械工件形变激光扫描识别中缺失点云配准方法  

Missing Point Cloud Registration Method in Laser Scanning Recognition of Mechanical Workpiece Deformation

作　　者：苏进 丁响林[1] 阚玉锦 SU Jin;DING Xiang-lin;KAN Yu-jin(School of Mechanical Engineering,Anhui Technical College of Mechanical and Electrical Engineering,Wuhu 241000,China)

机构地区：[1]安徽机电职业技术学院机械工程学院,安徽芜湖241000

出　　处：《辽东学院学报（自然科学版）》2023年第1期31-36,共6页Journal of Eastern Liaoning University:Natural Science Edition

基　　金：安徽省自然科学基金项目(YZ200917)。

摘　　要：为解决局部激光点云数据缺失导致识别结果准确度下降的问题,提出一种三维激光点云数据配准方法,并应用到机械工件形变识别中。利用三维激光扫描技术采集机械工件三维激光点云数据,经过聚类及逆变换处理后,完成机械工件激光扫描的点云配准;根据点云数据,获取机械结构在某点处的应变形态大小,通过计算机械结构的位移函数完成机械工件形变识别。实验结果表明,该方法可以有效配准机械工件的激光点云数据,提取的准确性高、耗时少,可有效实现精度较高的形变识别。In order to solve the problem that the lack of local laser point cloud data will lead to the decrease of recognition accuracy,a 3D laser point cloud data registration method was proposed and applied to the deformation recognition of mechanical workpieces.The 3D laser scanning technology was used to collect the 3D laser point cloud data of the mechanical workpiece.After clustering and inverse transformation,the point cloud registration of the laser scanning of the mechanical workpiece was completed.According to the point cloud data,the strain shape of the mechanical structure at a certain point was obtained,and the deformation identification of the mechanical workpiece was completed by calculating the displacement function of the mechanical structure.The experimental results showed that this method could effectively register the laser point cloud data of the mechanical workpiece,and accurately extracted the deformation point cloud data of the workpiece,and achieved high precision deformation recognition by spending the least time.

关 键 词：三维激光扫描 激光点云数据 机械工件 形变特征 缺失数据 数据配准 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]