Data fusion applied in a back warning system

During the last decades, many accidents occurred when reversing vehicles, our objective is to find a system that reduces the damages in case of backtrack (reducing accidents child injuries, avoiding hitting poles or walls, improving the appreciation of distance). Also this application aims at increasing the vigilance of the driver on the presence an object mostly nearby. The interest of this application is to combine several sources, since the sensor give us different kind of information, so with merging these data with the belief function theory we should decide which objects represent a danger for the vehicle in reverse.