The limitations of Global Navigation Satellite System (GNSS) in enclosed spaces have prompted the need for alternative methods for determining the position and navigation of Unmanned Aerial Vehicles (UAVs). One potential solution is to employ visual-based sensors and artificial intelligence (AI) algorithms. This study aims to explore the feasibility of a monocular vision-based positioning system for determining the position of a UAV by placing a sensor in a fixed position and feeding the image stream to an AI algorithm to generate the position and navigation solution. The results of this study could have practical applications, particularly inspection hangars where GNSS cannot be used. Automating the inspection process using UAVs could also help reduce the risk of human injury associated with elevated working spaces by improving the safety and efficiency of the inspection process.
Development of an Indoor Visual-Based Monocular Positioning System for Multirotor UAV
2023-10-26
466543 byte
Conference paper
Electronic Resource
English
Indoor positioning method based on monocular image stitching
European Patent Office | 2023
|System and method of multirotor dynamics based online scale estimation for monocular vision
European Patent Office | 2020
|