When encountering emergency conditions, the pilot may lose the help from the modern navigation systems and be forced to perceive the surrounding environment only through vision, which means the safety of manual landing depends heavily on human factors. To provide the pilot with additional information about the status of the aircraft, helping increase the safety during the landing procedure, ORB-SLAM2, a state-of-the-art SLAM algorithm, is used in this paper to estimate the 6DoF localization of the aircraft. To evaluate the performance for aircraft landing, this work used Unreal Engine to generate the simulation runway scene with built-in stereo cameras. The results show that for the landing scenario, ORB-SLAM2 can provide the pilot with an alarm message when the descending speed or the glide angle of the aircraft is inappropriate. Also, as a lightweight, real-time, and stand-alone features, ORB-SLAM2 could be easily applied to other smart and capability-limited unmanned aerial vehicles.