Abstract
Traditional industrial robot machining typically involves offline programming of machining trajectories, with part positioning accomplished through robotic tools. However, this method has limitations, particularly when dealing with complex workpieces and dynamic environments. To tackle this problem, this paper introduces a vision-based multi-view reconstruction approach for part positioning in industrial robot machining. Multiple cameras capture multi-view images of the part, reconstruct its 3D point cloud, and match it with the CAD model of the part to achieve high-precision positioning. This method effectively translates offline programmed robot trajectories to real-world scenarios, improving the flexibility and adaptability of robot machining. Experimental results confirm the high accuracy and robustness of the proposed vision-based reconstruction method, achieving a relative error of less than 2 mm for 85.7% of points, with an overall root mean square error of 1.09 mm.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call