ASSESSMENT OF ELECTRO-OPTICAL IMAGING TECHNOLOGY FOR UNMANNED AERIAL SYSTEM NAVIGATION IN A GPS-DENIED ENVIRONMENT

Abstract

Navigation systems of unmanned aircraft systems (UAS) are heavily dependent on the availability of Global Positioning Systems (GPS) or other Global Navigation Satellite Systems (GNSS). Although inertial navigation systems (INS) can provide position and velocity of an aircraft based on acceleration measurements, the information degrades over time and reduces the capability of the system. In a GPS-denied environment, a UAS must utilize alternative sensor sources for navigating. This thesis presents preliminary evaluation results on the usage of onboard down-looking electro-optical sensors and image matching techniques to assist in GPS-free navigation of aerial platforms. Following the presentation of the fundamental mathematics behind the proposed concept, the thesis analyzes the key results from three flight campaign experiments that use different sets of sensors to collect data. Each of the flight experiments explores different sensor setups, assesses a variety of image processing methods, looks at different terrain environments, and reveals limitations related to the proposed approach. In addition, an attempt to incorporate navigational aid solutions into a navigation system using a Kalman filter is demonstrated. The thesis concludes with recommendations for future research on developing an integrated navigation system that relies on inertial measurement unit data complemented by the positional fixes from the image-matching technique.