This paper proposes an artificial inspection auxiliary system based on AR technology to achieve high quality and efficiency in industrial inspection. The real scene is superimposed with virtual information via the AR terminal, and the inspection work is analyzed, guided, and decided, effectively assisting the artificial inspection work. To identify and match the equipment, SURF’s scale invariance and efficient identification performance are used. The high-precision indoor positioning of the tunnel 5G, the POI information of the object, the direction angle sensor in the AR intelligent terminal, and other data information are used to establish identification and matching based on the unique attributes of the target object, which greatly improves the accuracy of identification and matching. The system’s real-time data and decision-making function can reduce the average inspection time to 59% of that of traditional artificial inspection. The AR auxiliary inspection system-planned inspection route effectively avoids missed inspections, and overall inspection accuracy is close to 93%, effectively improving inspection quality and efficiency.