In order to improve the automation level of logistics industry and improve the work efficiency of material sorting、reduce labor costs, the control system for sorting manipulator based on two-dimensional code recognition is designed. This research selects Raspberry Pi as the main controller of the system, which can realize the material sorting function based on two-dimensional code by image processing with Open CV、two-dimensional code recognizing、inverse kinematics solving and driving the six-degree-of-freedom manipulator. In order to improve the accuracy of two-dimensional code recognition, the image is preprocessed by gray-scale、filtering、binarization and closing operation aiming at the wear and scratch problem of two-dimensional code. Aiming at the randomness of two-dimensional code position and lens distortion, the two-dimensional code was extracted by positioning algorithm and perspective correction before it is decoded according to the structure and coding rules of QR code. The inverse kinematics is calculated using simple geometric method and the speed control of multi-channel steering gear is added to the drive of manipulator for smooth and stable operation of manipulator. The experimental results show that the recognition rate of two-dimensional code is 100% and the manipulator can effectively complete the automatic sorting task, thus it has certain practical value.