Indoor route networks models are created for use in navigation. They may be built manually, but it is better to generate them automatically, based on the building floor plans. Research has been conducted in this field in many research centers. The authors undertook to develop their own methodology for generating navigation networks, using topological neighborhood relations and semantic data. The research project focuses on one floor in a building, which consists of rooms and an expanded corridor with an obstacle in the form of an open space between the floors. The first stage of the project consisted in the segmentation of the corridor space to improve its resolution. The objective of the conducted research was to select special points (five suggestions) for the segmentation. As a result, five different segmentations of the corridor space were obtained. The aim of the second stage was to automatically generate five navigation network models. The graphically presented results have been verified against the routes generated between the selected points in the building plan. A comparison of the results with other solutions shows that the routes generated in the presented methodology are more straight-line and less zigzagging.
展开▼
