Abstract
A vision based seam tracking system is designed for fillet weld in this paper. Laser structured light is used to detect the weld joint, and image-based visual control method is adopted. Due to the nonlinear relationship between the image feature change and the movement of adjusting mechanism, back propagation (BP) neural network is used to establish the vision model. Two independent controllers in two orthogonal directions are developed based on the vision model. To verify the effectiveness of the proposed seam tracking system, a series of experiments are well conducted. The results show that high tracking accuracy can be obtained using the proposed system and methods.
