Study on Point Cloud Processing Technology in Robotic Welding for Ship Sub-Assembly

To address the issue of insufficient automation in robotic welding for ship small assemblies, this study develops a welding technology based on point cloud processing. The research utilizes vision sensors to collect point cloud data of workpieces, and an improved point cloud processing algorithm is employed to extract weld seam information effectively. A novel real-time online calibration method for vision sensors is introduced, which precisely transforms the weld seam position from the vision sensor coordinate system to the robot coordinate system, thereby ensuring accuracy and continuity in the robotic welding process. Experimental results demonstrate that the proposed technology significantly enhances automation and weld quality in ship small assemblies, showing substantial industrial application potential.