Vision-based terrain following for an unmanned rotorcraft

Experiments from biology suggest that the sensing of image motion or optic flow in insects provides a means of determining the range to obstacles and terrain. When combined with a measure of ground speed from another sensor such as GPS, optic flow can be used to provide a measure of an aircraft’s height above terrain. We apply this principle to the control of height in a helicopter, leading to the first optic flow based terrain following system for an unmanned helicopter. Using feedback of the height estimated from optic flow ranging to the collective pitch control of the helicopter, it has been possible to maintain terrain clearance in flights of up to 2km. In this paper, we present flight test data demonstrating the succesfull application of this to an 80kg Yamaha RMAX unmanned helicopter and an 8kg electric helicopter. To complete this work, we have extended the optic flow Image Interpolation Algorithm (IA) to include an adaptive capability providing a greater dynamic range. The new algorithm, called the Iterative Image Interpolation Algorithm (IA) exhibits excellent robustness in an outdoor environment and makes it suitable for flight control in a real-world environment.