Secure and Intelligent Video Surveillance using Unmanned Aerial Vehicles

Abstract

In hard-to-reach areas, such as remote rural areas with limited infrastructure, a region with challenging terrain, or urban neighbourhoods that receive very little foot traffic or have a very small native population, the road and traffic conditions are often poorly recorded due to the lackluster Intelligent Transportation Systems in these areas. Nevertheless, using Unmanned Aerial Vehicles can be a viable solution to help survey and track conditions in these areas. In contrast to surveying, privacy is a very important issue and therefore in this paper, we present an Unmanned Aerial Vehicle equipped with a system that can help survey road conditions in remote areas without invading the privacy of the surrounding community. The system records and encrypts videos of the road and its conditions. Following the successful recording and encrypting of the video, the Unmanned Aerial Vehicle returns to the base station where it establishes secure communication with a base server and uploads the videos through an object detection application to report on the results of the surveyed area. The results were able to show the fundamental traffic anomalies from cars, trees, trucks, potholes, people, and animals with over 90% success rate for some of those objects compared to other pre-trained models.