A WLAN-based positioning system for indoor augmented reality services

Abstract

Almost all current location-based augmented reality (AR) services on smart phone are based on GPS and compass. However, in general GPS cannot work in the indoor environment. In this paper, we propose a WLAN-based positioning system for AR services using fingerprint architecture. Taking the feature of AR service into account, fingerprint architecture is improved in the system by the following methods: (1) Considering the power limitation of smart phone, we use k-d tree and the best bin first (BBF) algorithm to replace the traditional linear scanning in order to reduce computational cost; (2) We take advantage of the compass used in AR to improve the KNN algorithm; (3) Other resources such as pose estimation in AR are also utilized to improve the performance. Our result shows that the accuracy, computational cost and real-time performance of the proposed system can meet the demand of common AR services on smart phone.