A Method of Robotic Visual Tracking for a New Automatic Laser Welding Line

Abstract

An on-line visual tracking sensor system for the automation of laser welding processes in heavy industries is presented. A method of image process and features extraction for structured light image of laser welding has been developed. In order to set the range of interest, the Radon Transform is applied to determine the x and y coordinate in image coordinate system. For image de-noising, the median filtering and bi-level thresholding are used. Furthermore, the middle line of the laser stripe is extracted. Finally, a five step search method is carried out to examine the mid-point and gap width. A series of experiments show that the proposed method in this paper is effective, and its tracking detection precision is much better than traditional methods.