The marine economy plays a very important and positive role in the national economy and social development. Because the ocean surges have a great impact on the safety and efficiency of marine logistics hoisting machine. At present, the wave motion compensation is mostly based on ship-specific sensors such as IMU or MRU, and the cost is high. A combination of logos was used, and a vision-based robotic hoisting motion following control method was proposed. Firstly, the kinematics analysis and hand-eye calibration of the manipulator were carried out. By using visual marker, the target objects in complex environments can be detected and identified. The man direction and direct linear transfer algorithm was used to calculate the orientation and position of the marker. A combination marker resisting image noise and occlusion issues with pose confusion scheme was designed. A quaternion based Kalman filter was used for pose estimation of combined marker. Then by establishing the hoist system model, the pseudodifferential feedback compound control algorithm was used on the speed loop. Trajectory tracking performance was improved. Based on the double S-shaped curve motion planning algorithm, the target motion tracking and path replanning were performed, and the acceleration/deceleration method was used to generate motion trajectory. Based on the comprehensive experimental platform of marine automatic and intelligent lifting equipment which can simulate the offshore working environment, the visual-based hoisting robot arm winch lifting and following control method was studied. And static targets and dynamic target following experiments were performed, the feasibility of the hoisting lift control and the effectiveness of the follow-up control strategy were verified.