Abstract: Aiming at the problem of large errors and slow speeds in grasping and positioning of the board throwing station, a method of robot throwing board based on improved nine point calibration is proposed based on the six axis manipulator as the research platform.The mathematical morphology edge detection method is used to identify the product after the camera collects the photos, and the manipulator can automatically fine tune based on it. And the nonlinear optimization algorithm is used to calibrate the frame coordinates and camera end coordinates simultaneously.The experimental results show that the proposed method has high repetition positioning accuracy and can accurately and quickly locate, identify and grasp.