A Vision-Based GPS-Aided Quadcopter for Emergency Vehicle Assistance

A significant number of patients needing emergency treatment have died on their way to the hospital because of delays due to traffic jams and uncooperative motorists. The proposed work aims to build an automated quadcopter that follows a global positioning system (GPS) route which is visually guided by a camera that helps it to stay in the middle of the road. With this quadcopter assisting an ambulance, the delay caused by traffic jams can be reduced significantly. The project makes use of a simulated quadcopter and follows a proper route as a vehicle does by following waypoints extracted from Google Maps with the help of the Directions API. The waypoints are placed at significant turns in the route. To make sure the quadcopter stays in the middle of the road to avoid buildings and poles, a camera placed on the quadcopter streams a live feed to a computer that runs a road detection program using a fully convolutional networks. The program provides correction feedback signals to the quadcopter.