Indoor-outdoor navigation without beacons: Compensating smartphone AR positioning errors with 3D pedestrian network

Abstract

Despite the extensive use of positioning and navigation in outdoor space, indoor positioning and navigation systems, essential for intelligent building and smart city services, are unsatisfactory in either performance or price, sometimes in both. This paper analyzes and compares the performances and prices of existing indoor positioning technologies that are categorized into a few classes according to their spatial sensing and referencing methods. Based on the previous work in walkability, this paper proposes a novel Walkability network-based Augmented Reality (WaNAR) method using smartphones with AR positioning function for positioning and navigation. In WaNAR, drifting of the AR positioning signals are corrected continuously by the ground-truth 3D indoor/outdoor walkability network (e.g., nobody is supposed to walk through a wall) in a 3D model. The error at the vertical axis of the walking direction is corrected continuously and that of the walking direction is compensated at every turn. WaNAR can be used in both indoor and outdoor navigation, its performance and price are proved to be largely improved compared to existing technologies. The only investment for a typical building is a 3D drawing of indoor walkable space in a few staff-hours. WaNAR has broad application prospects at various positioning and navigation scenarios.