Seamless Navigation in Indoor and Outdoor based on 3D Spaces

Abstract

Contemporary living environments are getting more and more complex, combining structures in indoor and outdoor. These structures can be broadly subdivided into entirely bounded (indoor), partially bounded (semi-bounded), and unbounded (outdoor). Normally, agents (e.g., pedestrians) wish to have seamless navigation when seamlessly transferring from one kind of environment to another and back. Seamless here means human operations or interventions are not required when transitions happen between different navigation modes and environments.

In the past decades, seamless navigation has attracted a lot of attention and many approaches have been reported in the literature or made available as commercial applications. Most of them rely on integrating indoor navigation networks to outdoor road/street-based networks via specific structures (e.g., anchor node). Yet, navigation paths from such approaches are not seamless, which can be partly explained by the fact that navigation networks in indoor and outdoor used for integration are currently built differently; due to semi-bounded spaces have unclear definitions thus such spaces are often omitted from navigation maps.

Therefore, this thesis developed new theories, models, and approaches to support seamless navigation in all kinds of environments, which include a novel generic space definition framework, a unified 3D space-based navigation model, approaches for automatically reconstructing 3D spaces, and two new path options (MTC-path and NSI-path). This research brings in four contributions to seamless navigation: (i) living environments (spaces) are systematically categorised and defined as indoor (I-space), semi-indoor (sI-space), semi-outdoor (sO-space), and outdoor (O-space); (ii) all types of spaces are uniformly modelled and managed as 3D spaces, thereby sharing the same approaches for deriving navigation networks; (iii) sI-spaces and sO-spaces are included in navigation maps, which allows developing new path options based on them; and (iv) on the basis of 3D spaces, vertical constraints are considered in navigation path planning.