Study of robot seam tracking system with laser vision

Welding is developing to automatization and robotization, and robot welding is an important development direction. A robot real-time seam tracking system based on laser vision is researched, which includes laser vision sensing and seam image acquisition, seam images recognition and extraction, system coordinates calibration and transformation, communication between PC and robot, control for robot seam tracking, and robot seam tracking experiments. A way of image filtering is improved and four methods to recognize and extract seam images are studied. Lap-joint seam tracking experiments are performed to test the seam tracking system. Lap-joint groove experiment results show, corner detecting image recognition is satisfied with real-time seam tracking and has good precision and stability, and the whole system studied in the paper is satisfying for robot real-time seam tracking.