POSITION ESTIMATION OF AERIAL VEHICLE BASED ON A VISION AIDED NAVIGATION SYSTEM

Automatic pose estimation remains to be one of the challenging problems in photogrammetry and computer vision. When other systems (e.g. GPS and INS) are inadequate, and real time precise positioning is required, image is a very high quality device for localization of vehicles. In this paper, we address a method for Vision Based Navigation (VBN) of Aerial Vehicles which is to determine precise position and attitude of the vehicle in real time. Pose Estimation is achieved by registering single sensed image frames to stored reference database using salient features and descriptors. The reference database includes previously collected satellite imageries and 3D terrain maps that have been precisely aligned to geo-coordinates. The pose estimation strategy has been tested on a number of different images and experimental results proved the feasibility and robustness of our approach. 1