Light intensity analysis of a passive visual sensing system in GTAW

Abstract

In manual welding, an experienced welder acquires 80% of the welding process information by vision. Visual sensing has the advantage of supplying abundant information without contact with the workpiece or welding circuit. Visual sensors have been widely used in the field of welding. According to optical imaging principles, the object imaging process depends on the shape, size, position, and angle of the light source; the shape of the object; the reflectance characteristic of the object’s surface; and the position, angle, and inner parameters of the camera. In order better to utilize the image information of a visual sensor and to ascertain the weld shape and size, arc light intensity is calculated from the arc length and the welding current. The visual sensing system is analyzed from the point of the view of light intensity. The function of adjusting the system to the light intensity is formulated in this paper. A relationship between the gray image, arc length, and welding current can be obtained. Welding experiments were implemented. Thus, we can recover the shape and height of the weld pool by SFS (shape from shading) algorithms from the weld pool image.