In many scenarios, people have to walk through unfamiliar indoor spaces such as large airports, office buildings, commercial centers, etc. As a result, indoor navigation is of realistic importance and great potential. Existing indoor space models for indoor navigation assume that relevant indoor space information is already available and precise in the model-specific format(s). However, such information, e.g., indoor topology that is indispensable to indoor navigation, is only implicitly (and even imprecisely) hidden in industry standards like the Industry Foundation Classes (IFC) that describe building projects. This paper is motivated to bridge the apparent gap between industry standards and indoor navigation. In particular, we propose an effective method to construct indoor topology by carefully processing IFC files. We also refine an existing method that decomposes large and/or irregular indoor partitions, which helps speed up routing in indoor navigation. Furthermore, we design an algorithm that computes indoor distances involving concave partitions. We conduct extensive experiments to evaluate our proposals. The experimental results demonstrate that our proposals provide effective processing of IFC files and efficient indoor navigation.
展开▼
