Visual positioning system for automated indoor/outdoor navigation

This paper proposes a proof-of-concept for a novel automated indoor/outdoor navigation system. Our proposed method shall enable an object/user equipped to be able to navigate through closed environments using an automatically generated Spatial Map Graph (SMG) with the aid of pre-placed visual markers. The system is robust to dynamically changing complex environments, through adaptive reconfigurations in the SMG during execution. We show that it is possible to find optimal routes among several interconnected paths without relying on an external positioning system. Points of interest in a passive environment are detected from real time first-person-view input, which is then processed to pinpoint the current location, and extended further for navigation. Through this paper we explore the advantages and the challenges faced in implementing such a navigation system in a practical scenario. The way forward and future work has also been discussed.