Flood Rescue Operations Using Artificially Intelligent UAVs

This research work proposes the integration of artificial intelligence (AI) in the field of disaster relief operations. This work particularly stresses disaster relief operations involving water such as floods and tsunamis. The primary emphasis of this study is on human and water detection. Human detection will aid the process of search and rescue operations, while water detection will aid the artificially intelligent unmanned aerial vehicle (UAV) to hover only over the water body and avoid hovering over dry land. A dataset of approximately 16,200 frames at the frame rate of 15 frames per second containing humans stranded in the water body was collected and upon initial testing, the study revealed an average accuracy of 70%. The findings indicate that the UAV is capable of performing search and rescue operations with tremendous accuracy and that too in realtime.