Stacking Objects Picking System Based on Structured Light

This paper proposes a method for stacking objects picking system that combines two-dimensional vision and three-dimensional vision. Structured light method is used to acquire three-dimensional information. Based on the information, the region of interest is extracted. A Generalized Hough Transform method is then used to match object and estimate the pose of object. Finally, the pose of the object is sent to the manipulator for automatic picking. The method proposed in this paper is very effective for the capture of stacking flat objects. The combination of two-dimensional and three-dimensional methods improves the efficiency of the algorithm. The experiments show that the system has high performance.