Tightly Coupled GNSS and Vision Navigation for Unmanned Air Vehicle Applications

This paper explores the unique benefits that can be obtained from a tight integration of a GNSS sensor and a forward-looking vision sensor. The motivation of this research is the belief that both GNSS and vision will be integral features of future UAV avionics architectures, GNSS for basic aircraft navigation and vision for obstacle-aircraft collision avoidance. The paper will show that utilising basic single-antenna GNSS measurements and observables, along with aircraft information derived from optical flow techniques creates unique synergies. Results of the accuracy of attitude estimates will be presented, based a comprehensive Matlab® Simulink® model which re-creates an optical flow stream based on the flight of an aircraft. This paper establishes the viability of this novel integrated GNSS/Vision approach for use as the complete UAV sensor package, or as a backup sensor for an inertial navigation system