Abstract: To address the issues of excessive manual intervention， difficulty in posture adjustment， and low precision during the hoisting process of prefabricated building wall panels， a collaborative control method integrating intelligent hooks， self-balancing lifting beams， and machine vision is proposed. An intelligent hook coupling a magnetic adsorption positioning mechanism with a two-stage gear reduction mechanism is designed to achieve automatic hooking and unhooking of lifting rings. A self-balancing adjustable three-point lifting beam based on ball screws is developed， employing a posture sensor-PID closed-loop control strategy to enable continuous tilt angle adjustment within the range of 0.5° to 15°. A binocular vision-embedded fusion recognition system is constructed， utilizing a geometric contour matching algorithm to achieve 3D feature reconstruction of wall panels. Through 200 sample tests， the type recognition accuracy reached 98.2%. The system integrates LoRa wireless communication and a modular architecture. Engineering tests demonstrate that the entire hoisting process for typical wall panels is reduced to 62.4% of the time required by traditional methods， providing an efficient and highly reliable solution for automation in prefabricated building construction.