An Integrated Position and Attitude Determination System to Support Real-Time, Mobile, Augmented Reality Applications

Augmented reality (AR) technologies enable digitally stored information (virtual objects) to be overlaid graphically on views of the real world. As such, they are able to significantly enhance decision-making and operational efficiency in complex environments. AR technologies typically comprise a fusion of positioning and attitude sensors with visualisation capability and an information processing system. The decreasing size and cost of visualisation and positioning hardware and the increasing portable processing power of laptop and handheld computers now offer enormous potential for the development of intelligent solutions based around realtime, mobile AR technologies. For any application built around AR technologies, its effectiveness lies in the accuracy to which the virtual objects can be aligned with views of the real world. For many of these applications, this is directly a function of the accuracy to which the position and orientation of the operation platform can be determined. This paper presents an integrated positioning system that combines an array of dual frequency GPS receivers, a fibre optic gyroscope and vehicle odometer within a centralized Kalman filter. It assesses the accuracy of the filter outputs of position and attitude as appropriate to supporting realtime, mobile AR applications. The design and testing of an AR prototype that combines the Kalman filter state with real-time imagery containing augmented objects will also be presented. Finally, approaches adopted to tune the filter and reduce inherent sensor noise, as well as results from a case study undertaken within the land mobile environment will be described.