The welding process is the most time and cost consuming process and the quality of welding is also an important issue in shipbuilding. In order to improve these problems, welding automation robots are widely used. For welding automation, it is necessa...
The welding process is the most time and cost consuming process and the quality of welding is also an important issue in shipbuilding. In order to improve these problems, welding automation robots are widely used. For welding automation, it is necessary to make the welding line of the design data into the robot coordinate system so that the robot can work. In this study, we propose a method to generate welding paths using a lightweight model and a point cloud. Welding lines are extracted using the lightweight model already built for digital collaboration in shipyards. And the work targets are recognized using the point cloud data obtained from the depth camera. Through the grid construction, noise is removed from the measured data, and stiffeners and boundary data are generated. Weld lines are mapped to work targets using ICP (iterative closest point). This proposed method has been tested on several real models to demonstrate the feasibility.