Swarming Drone Sensors

Abstract

In the last decades, the presence of UAVs has increased widely in the military world, as they are able of monitoring conflict areas without endangering human lives. Many of these UAVs have the disadvantage of being quite big and expensive; therefore, the trend now is to use lots of smaller and cheaper drones which make it possible for the system to continue working even if a couple of drones get lost or are unable to contribute. In this project it has been designed a robust system that using a fixed number of drones with single cameras on them delivers a good resolution picture comparing with the ones that are obtained from expensive systems. Given a certain mission area, first a task allocation algorithm assigns some tasks or positions in the mission area to each UAV, in a way that the information collected by the images is maximized. After that, an image mosaicing algorithm will process those images in order to return the final mosaic. The whole thesis has been developed in a simulation environment in Matlab. The results show that the proposed algorithms guarantee that the complete mission area will be covered by the UAVs in the shortest possible time. In addition, the obtained final mosaic represents perfectly the considered mission area when an adequate overlapping area is considered. Finally, the system has been proven to be resilient if a UAV is unable to contribute for some reason.