The stacking robot has complex information about the types and positions of materials stacked, numerous grasping targets and easy visual overlap with the robot structure, which makes it difficult for a single PID control technology to accurately identify the connection form of the robot arm servo control structure, resulting in a low success rate of the robot in different stacking operation scenarios. Therefore, the object stacking technology of vision guided controllable metamorphic stacking robot is proposed. The material image is captured by high-precision camera, and then analyzed by machine vision software, so as to accurately identify the material type and position information and complete the visual guidance of the manipulator. Analyze the connection form of the servo control structure of the vision guided manipulator, reduce the visual overlap, build the dynamic model of the stacking operation, and then combine the PID feedback controller to achieve the precise control of the robot"s stacking operation behavior. Use the target stack sample points to build the pose point data set, and implement enhancement processing for the data objects in it. Evaluate the grab priority of stacking targets, and judge the auxiliary stacking relationship between the determined target stacking points. On this basis, the stacking trajectory of the robot is planned to realize the design of the target stacking technology of the controllable metamorphic stacking robot. The experimental results show that the above stacking techniques achieve a success rate of 100% in single target and simple stacking scenarios, and a success rate of over 90% in complex stacking scenarios.