Vision-Based Unmanned Aerial Vehicle Navigation Using Geo-Referenced Information

• DOI: 10.1155/2009/387308
• Publication Date: 2009-08-11T11:00:00Z
• Authors: Gianpaolo Conte; Patrick Doherty
• Keywords: Odometry; Visual Odometry; Position (finance); Navigation System

This paper investigates the possibility of augmenting an Unmanned Aerial Vehicle (UAV) navigation system with a passive video camera in order to cope long-term GPS outages. The proposes vision-based architecture which combines inertial sensors, visual odometry, and registration on-board geo-referenced aerial image. vision-aided developed is capable providing high-rate drift-free state estimation for UAV autonomous without system. Due use image-to-map absolute position calculation, performance depends on structural characteristics terrain. Experimental evaluation approach based offline flight data provided. In addition proposed has been implemented experimental helicopter platform tested during flights.