Machine vision and scientific imaging for autonomous air vehicles (UAV).

This thesis outlines the necessary requirements to determine an Unmanned Aerial Vehicles (UAV’s) pose relative to a lead aircraft or target, thus enabling a UAV to successfully follow a lead aircraft or target. The use of Machine Vision for Autonomous navigation has been investigated and two flight scenarios were chosen for analysis. Firstly, following a manoeuvring lead aircraft, and secondly, maintaining a steady heading behind a target/lead aircraft (as would be required for in-flight refuelling). In addition, the author has performed a literature review of current research in this field which is significantly dominated by eventual military requirements in order to improve UAV endurance. In addition, experimental work towards developing a passive vision based navigation system has been undertaken. It is hoped that after further research and development this will lead to an eventual flight trial using the flight dynamics department’s UAV’s. The experimental work has been performed using both equipment and software already available within the department and furthermore, it has enabled an analysis of the department’s currently available capabilities for passive visual navigation to be undertaken. Key points for further work have been outlined for the future advancement of the visual navigation project.