Human Detection in Drone Images Using YOLO for Search-and-Rescue Operations

Today, unmanned aerial vehicles, more commonly known as drones, can be equipped with high-resolution cameras and embedded GPUs powerful enough to provide effective and efficient aid to Search-and-Rescue (SAR) operations in remote and hostile environments. Locating victims, who may be unconscious or injured, as quickly as possible is critical to improving their chance of survival. Therefore, using drones as flying machines for computer vision can increase the detection rate while reducing rescue time. In this paper, we present the results of an experimental evaluation in which we used the latest, lightweight version of the YOLO detection algorithm, namely YOLOv5, to detect humans in danger using two new benchmark datasets specifically designed for SAR with drones. The results obtained are encouraging, as they are competitive with respect to the state-of-the-art in terms of detection accuracy, but with much faster detection time.