Recent developments on welding image processing and features extraction

Robots have been introduced to the welding processes to describe the welding image processing, profiles extraction, weld joint recognition and feature extraction. Welding robots guided by sensors can adjust the track, pose of torches and parameters automatically in welding process. Various intelligent algorithms such as fuzzy logic, template matching and neural networks have been developed for the robust recognition and extraction of the weld joint. The area near to seams of images only need to be processed to reduce the computational cost for image processing. Structured light images of weldment are usually strongly contrasted, where the images are naturally segmented to detected stripes roughly by threshod. Image filtering and de-noising is one of the important issues in image processing research. The edge detection method has high efficiency with low computational complexity because of the concision of gradient operators.