The visually assisted landing system lets the UAV uses its camera to identify landmarks, adjust speed and direction accordingly, and navigate to a smooth landing procedure.
The system automatically picks out features on the ground, specifically the runway marking, then tracks the way these features change from frame to frame and builds up a 2D model of what they represent, so that relative motion can be calculated. Combined with GPS, the system can land anywhere without an elaborate runway preparation.
The landing system has the advantage that it’s looking at a tidy, structured scene (runway markings) rather than chaotic natural scenery. Of course, runway markings are designed to be as visible and unambiguous as possible; the system simply takes advantage of this. With the aid of accelerated computer vision algorithms, it can run at thirty frames per second, fast enough to cope with the approach rate of a landing UAV.
In collaboration with Formal Methods Group  of Politecnico di Torino – Department of Computer Science