基于DoN-LMedS的焊道点云分割方法  

Adaptive Segmentation Method for Weld Seam Point Cloud Based on Normal Features and LMedS

作　　者：杨宏韬[1] 李鑫 程云龙 李秀兰[1] 张鹏[2] 白昊天 YANG Hongtao;LI Xin;CHENG Yunlong;LI Xiulan;ZHANG Peng;BAI Haotian(School of Electrical and Electronic Engineering,Changchun University of Technology,Changchun 130012,China;Petroleum Natural Gas Pipeline Engineering Company Limited Tianjin Branch,Tianjin 300457,China)

机构地区：[1]长春工业大学电气与电子工程学院,长春130012 [2]中国石油天然气管道工程有限公司天津分公司,天津300457

出　　处：《组合机床与自动化加工技术》2025年第3期209-213,218,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然基金项目(62106023);吉林省科技发展计划项目(20240404055ZP);吉林省教育厅科学技术研究项目(JJKH20240856KJ)。

摘　　要：针对由三维扫描设备直接采集的焊道点云与母材点云融合,对后续焊道处理工艺造成干扰且传统点云分割算法对焊道点云分割效果不佳的问题,提出了一种基于DoN-LMedS的焊道点云分割方法。首先,利用最小平方中值(LMedS)拟合焊板平面点云,提取焊道坡顶点云;其次,计算点云的法线微分算子(DoN)得到曲率较大的焊道根部点云并依据边缘高度差计算二次LMedS分割阈值,得到焊道根部点云;最后,将两次分割后的结果进行点云拼接获得完整的焊道点云。实验结果表明,该方法能够正确提取焊道点云,分割精确率、召回率、分数分别能够达到97.91%、98.38%、98.32%,且对不同形状焊道的提取具有较好的鲁棒性。在相同的实验条件下,所提算法的计算复杂度更低,具有更快的焊道点云分割速度。To address the interference in weld processing caused by the fusion of weld seam and base material point clouds collected by 3D scanning,and the poor performance of traditional segmentation algorithms,a DoN-LMedS-based segmentation method is proposed.Firstly,the minimum square median(LMedS)is used to fit the plane point cloud of the weld plate and extract the weld seam crest point cloud.Secondly,the curvature of the weld seam root point cloud with large curvature is calculated using the point cloud normal differential operator(DoN),and the secondary LMedS segmentation threshold is calculated based on the edge height difference to obtain the weld seam root point cloud.Finally,the results of the two segmentation are merged to obtain the complete weld seam point cloud.The experimental results demonstrate that the method is capable of accurately extracting weld seam point clouds,the segmentation accuracy,recall rate and score can reach 97.91%,98.38%and 98.32%respectively.Furthermore,it exhibits favorable robustness in extracting weld seams of various shapes.Under the same experimental conditions,the computational complexity of this algorithm is lower,resulting in faster weld point cloud segmentation speed.

关 键 词：三维点云 焊道分割 最小平方中值 法线微分算子 

分 类 号：TH165[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]