Virtual machine vision in computer aided robotics
ETFA'09 Proceedings of the 14th IEEE international conference on Emerging technologies & factory automation
| Hi-index | 0.00 |
Vision-based seam tracking without auxiliary light is a popular topic concerning the intelligent welding robot. The Gordian Knot of the issue is image processing with much time spending in the welding course. First, according to the principle of paraboloid of revolution reflecting, this paper puts forward laying a paraboloidal mirror in specifically position above the tip of welding wire. A part of arc light is changed into parallel one by the mirror. The reflected light shines on the area where the CCD camera can see. So the visual seam is seen long and the arc light jamming is weakened in the CCD camera. Secondly, it will be discussed that the luminous flux of the area is assured. Then the Roberts algorithm for the processing image will be discussed in the aspect of the positiveness and negativeness of the gradient element, a rule that judged seam was brought forward. The rule increases the speed of seam tracking. A welding robot system with single camera fixed on the weld torch end-effector has been established for the robot to identify the welded seam by the reflected light. The simulation and results of real application have proved the validity of the method.