Multi-target self-adaptive angle grabbing method of visual servo mechanical arm

A global camera system and a hand-eye camera system are combined, the multi-target self-adaptive angle grabbing method of the visual servo mechanical arm is disclosed, the position information of a grabbing target is obtained mainly according to a global camera, and the mechanical arm is moved to the initial position of a grabbing task in cooperation with the hand-eye camera; it is guaranteed that the target object is within the visual field range of the hand-eye camera, then the mechanical arm is controlled to adjust the tail end angle according to the hand-eye camera through the MPC algorithm, and a grabbing task is conducted. According to the visual servo control method for the mechanical arm, the mechanical arm can accurately grab a large number of objects of various types in a self-adaptive angle mode under the complex background environment, and the flexibility, adaptability and reliability of the visual servo mechanical arm are improved. 本发明将全局摄像机系统与手眼相机系统结合，公开了一种视觉伺服机械臂的多目标自适应角度抓取方法，其主要依据全局相机得到抓取目标的位置