基于面结构光的焊缝坡口特征提取  

Weld Groove Feature Extraction Method Based on Surface Structured Light

作　　者：余明岭 龚烨飞 刘霞[1] 刘继承 谷心浩 谢雨欣 YU Mingling;GONG Yefei;LIU Xia;LIU Jicheng;GU Xinhao;XIE Yuxin(School of Physics and Electronic Engineering,Northeast Petroleum University,Daqing 163318,China;School of Electrical and Automation Engineering,Changshu Institute of Technology,Changshu 215500,China)

机构地区：[1]东北石油大学物理与电子工程学院,大庆163318 [2]常熟理工学院电气与自动化工程学院,常熟215500 [3]中国矿业大学信息与控制工程学院,徐州221116

出　　处：《组合机床与自动化加工技术》2023年第12期17-20,共4页Modular Machine Tool & Automatic Manufacturing Technique

摘　　要：针对机器人视觉导引定位焊缝坡口特征应用需求,提出了一种基于面结构光视觉传感器扫描与点云切片投影的坡口特征提取方法。首先,使用面结构光视觉传感器对焊接工件进行扫描,以获得焊缝坡口点云数据;其次,通过点云数据的最小有向包围盒得到等间距排列的切平面,将平面内数据点由三维空间投影至二维图像;最后,通过最大距离法粗提取与拟合直线相交法精提取相结合方法获得焊缝坡口特征点。实验结果表明,该方法能够准确地提取出焊缝坡口特征,识别误差小于1.0 mm,且在成像效果不佳的工况下具有较好的鲁棒性。A feature extraction method for weld seam based on surface structured light vision sensor scanning and point cloud slicing projection was proposed to extract weld seam groove feature in vision-based robotic welding.Firstly,weld seam groove point cloud data were obtained through surface structured light vision sensor scanning.Then equidistant cutting planes were established based on the oriented bounding box of the point cloud data,and the points within cutting plane were projected from a 3D space onto 2D image.In the final step,weld seam groove feature points are obtained through a combination of the maximum distance method and the intersection of fitted lines method.The test result showed that the proposed methodology could accurately extract weld seam groove features with a recognition accuracy of less than 1.0 mm,and has strong robustness in undesirable situations.

关 键 词：面结构光 焊缝特征提取 点云切片 图像处理 

分 类 号：TH162[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]