Space classification for indoor pedestrian navigation with morphological and functional characteristics

Peng Tian,Yunjia Wang,Yong Wang,Yuan Yang
DOI: https://doi.org/10.1049/ell2.13190
2024-04-27
Electronics Letters
Abstract:The letter proposes a method for classifying indoor navigation spaces based on their morphological and functional characteristics, improving the representation of complex indoor structures in topological network models. This approach classifies indoor navigation spaces into corridors or open spaces, which are suitable to be represented by median axis models or visual graphs respectively, addressing the limitations of width‐based parameters. Indoor map experiments confirmed the method's effectiveness in identifying various types of corridors and open spaces in complex indoor settings. Indoor navigation networks serve as the foundation for indoor pedestrian navigation services. However, current graph‐based topological network models cannot easily represent complex indoor structures in a manner consistent with user behaviour and cognitive patterns, and thus cannot support fine‐grained indoor navigation analysis services. To enhance the rationality of the model, we considered the influences of morphological and functional characteristics of indoor spaces on pedestrian movement and proposed a method for classifying navigation spaces. This method, based on convex space segmentation, defines morphological and functional characteristic functions, classifying indoor navigation spaces as either corridors or open spaces, which suitable for representing as median axis models or visual graphs, respectively. These categories supported the creation of topological network models for route planning analysis. Indoor map experiments confirmed the effectiveness of our method in identifying various types of corridors and open spaces, that addressing the limitations of width‐based parameters, even in complex indoor settings.
engineering, electrical & electronic
What problem does this paper attempt to address?