Vision Based UAV Formation Sense and Avoid System

With the increasing use of airspace and the wide application of unmanned aerial vehicles (UAV), the Sense And Avoid (SAA) problem has attracted more and more attention. This paper presents a solution for the SAA of UAV formation equipped with optical sensors. We introduce the stereo vision theory combined with the UAV formation to calculate the position of the targets. The tracking of target is resolved by adopting multi-model (MM) theory so as to solve the problem of the target maneuver. Also during the avoidance, we introduce ellipsoid security region as the safety region and use the flight angle to re-configure the formation’s flight. The simulation results show the effectiveness of this method. In the end, we described our test platform which is used to test the vision based SAA algorithm. Keywords—UAV formation; Vision based; Sense and avoid;