Proposed Integrated Drone Navigation and Autonomous Flight System for Aerial Inspection and Surveillance of Infrastructure Objects

Abstract

This paper describes the successful development of an integrated drone navigation system for aerial inspection and surveillance of infrastructure objects that features outbound and inbound GPS-based autonomous flight, machine learning of starting point and end point images of the destination structure, and a surface image gyro that detects images between the starting point and the end point. Traditional drone navigation research has focused on discrete autonomous flight technologies such as GPS navigation and non-GPS video SLAM. Thinking in terms of actual autonomous flights of drones, this study adopts a comprehensive approach that encompasses the entire trip to a destination from the departure of the drone up to its return.