Autonomous Surveillance Drone

Gaurav Shetty, Vraj Shah, Parva Juthani, Unik Lokhande

Abstract: Drones are aircraft without pilot, crew or passengers on board. They are formally known as unmanned aerial vehicles which can be controlled remotely by humans using a controller or even be completely autonomous. They can be used for various purposes like aerial photography, delivering items, reconnaissance and other civil and military applications. Our project aims to develop a drone which is completely autonomous and used for surveillance purposes. The user just needs to input the GPS coordinates of the path where it wants the drone to go, and then the drone automatically generates the path and starts surveillance. We have also incorporated computer vision to detect objects, human and other aircraft using the state of the art YOLOV4 algorithm. Successfully tested all systems with positive results.

Keywords: Optical flow, YOLOV4, INS, IRS, Computer Vision, PID