Vision-based weld seam tracking in gas metal arc welding

Abstract

A low-cost visual sensing system is developed to realize weld seam tracking in gas metal arc welding (GMAW). The system consists of a commercial CCD camera, narrow-band composite filter lens, an image capturing card, an industrial computer, a welding control unit, a GMAW power source, and a worktable. Images of root gap and its vicinity are captured in the GMAW welding process by the system. The captured images are processed by an algorithm on the basis of the analysis of gray characteristics of the root gap to get the offsetting information between torch and root gap centerline. The offsetting information is then used to realize weld seam tracking in the GMAW process. Welding seam tracking experiment is conducted by a simple proportional (P) controller. The results show that tracking error is basically less than ±0.5 mm.