A high performance altitude navigation system for small rotorcraft unmanned aircraft

An altitude navigation system is proposed for the small rotary wing unmanned aircraft to realize stable landing control. Based on the speeded up robust feature algorithm, the stereo vision system can get the feature matching points from the complex environment easily. With the consideration of tilt angle of SRUA system, the corresponding altitude estimation value can be generated by the parallax of ground points, and the stereo vision system can offer accurate altitude information without specialized landing marks. Meanwhile, with the analysis of the characteristic of the on-board altitude sensors, an adaptive weighted average filter algorithm has been developed to adjust the weights of onboard altitude sensors on line. Therefore, the navigation system can generate high performance altitude information even if some onboard sensors are invalid in the autonomous landing process. A series of static tests and autonomous landing tests verify the effectiveness of the altitude navigation system.