A Vision Driven Automatic Assembly Unit
The development of a flexible assembly unit is one of the de- manding tasks in industrial manufacturing. A higher degree of flexibility is mostly payed by an increasing complexity of the involved hardware. In this paper we present a three-step concept for a vision driven automatic assembly unit. These three steps are robust bin-picking to isolate objects from a pile of unorganized parts, exact pose determination to enable industrial mounting and visual inspection of the final assembling. For robust bin-picking we present a new structured light approach. Experiments show the robust and accurate behavior of the proposed algorithm and motivate the implementation in an industrial system. For exact pose determination, the second step, a pose estimation based on a modified view based approach, followed by a model based refinement is proposed. Initial experiments promise a fast and exact pose determination.
KeywordsPlane Detection Stereoscopic Image Plane Representation Assembly Unit Vacuum Sucker
Unable to display preview. Download preview PDF.
- 8.Krisnawan Rahardja and Akio Kosaka. Vision-Based Bin-Picking: Recognition and Localization of Multiple Comlex Objects Using Simple Visual Cues. In Proceedings of the International Conference on Intelligent Robotics and Systems, Osaka, Japan, November 1996. IEEE / RSJ.Google Scholar
- 9.Martin Rutishauser and Frank Ade. From Vision to Action: Grasping Unmodeled Objects from a Heap. In Intelligent Robots and Computer Vision XIV, SPIE’s Photonic East Symposium, volume 2588, pages 375–386. The international Society for Optical Engineering, October 1995.Google Scholar
- 12.Marjan Trobina and Aleš Leonardis. Grasping Arbitrarily Shaped 3-D Objects from a Pile. In International Conference on Robotics and Automation, volume 1, pages 241–246. IEEE, May 1995.Google Scholar
- 15.P. Wunsch and G. Hirzinger. Registration of CAD-Models to Images by Iterative Inverse Perspective Matching. In Proc. International Conference on Pattern Recognition, pages 78–83, 1996.Google Scholar
- 16.Billibon H. Yoshimi and Peter Allen. Closed-Loop Visual Grasping and Manipulation. In Proceedings of Image Understanding Workshop, pages 1353–1360, Palm Springs, California, February 1996.Google Scholar