基于点云的曲线焊缝识别方法研究

Research on curve weld identification method based on point cloud

三维曲线焊缝是焊接工艺中较为常见的一种焊缝,其具有复杂多变和曲线不规则等特点,使得对其识别难度大、精度低。针对自动化焊接过程中的三维曲线焊缝提出了一种基于点云的识别方法,该方法具有效率高、识别精度高、应用范围广以及可靠性强等优点。其原理是通过使用三维面结构光扫描仪获取工件表面的原始点云数据,通过使用PCL点云库进行降采样、分割、边界提取和曲线拟合等,实现焊缝边界的精确的提取,并拟合出焊枪末端的实际运行轨迹。

Three-dimensional curve weld is a common weld in welding process.It has the characteristics of complex,variable and irregular curve,which makes it difficult to identify.This paper proposes a point cloud-based identification method for three-dimensional curve weld in automatic welding process,which has the advantages of high efficiency,high identification accuracy,wide application range and high reliability.The principle is to obtain the original point cloud data of the workpiece surface by using a three-dimensional structural light scanner,realize the precise extraction of the weld boundary by using the PCL point cloud library for sampling,segmentation,boundary extraction and curve fitting,and fit the actual running track of the end of the welding gun.