D. Aiteanu and A. Gräser (Germany)
Process modelling, surface reconstruction, virtual and augmented reality
The research and development of a computer aided training system for the manual welding represents the framework of this paper. The system is based on the video-see-through augmented reality technique and uses a special designed welding helmet which integrates two high-dynamic-range cameras and a stereoscopic head worn-display. By processing the image information, the spatial position of the welding torch with respect to the work piece is estimated. Hence, relevant technological parameters like the inclination of the torch and its moving speed are calculated. The core of the paper consists in generation of a virtual welding seam, aiming the estimation of relevant properties of the real welding seam in real time. A dynamic model of the real seam is defined, in order to provide instant and accurate quantitative estimation. With the help of virtual three-dimensional elements registered with the real scene, the computer indicates corrections to be made in moving the torch and gives a quality feedback on the produced seam, indicating among others the seam’s thickness and cross section’s area. Providing real and simulated training programs, the system is able to increase the performance of the manual welder, to accelerate the training and to reduce the costs.
Important Links:
Go Back
