A data model for organizing relative semantics as images to support pedestrian navigation computations

Abstract

Pedestrian navigation systems often use the relative semantics of pedestrians and their environments to provide navigation guidance. Relative semantics include spatial and visual semantics. However, most navigation data models are based on an absolute reference frame and do not support the organization of relative semantics. To address this deficiency, we propose a pedestrian navigation data model based on relative semantic images that organizes the relative semantics of landmarks and environments directly in the image channels. Using geographic data for a university campus, we compared the data file size, data access time, and memory usage to confirm that the proposed approach outperforms the geodatabase approach in storing and accessing the relative semantic data. Two examples, self‐localization and route guidance, demonstrate the feasibility of the proposed data model. This model can support fast pedestrian navigation on mobile devices in small and medium‐sized areas.

Personalized landmark adaptive visualization method for pedestrian navigation maps: Considering user familiarity

“…Landmarks are essential elements in cognitive maps that can aid pedestrian navigation and provide cues to enhance spatial orientation and route guidance (May, Ross, Bayer, & Tarkiainen, 2003; Millonig & Schechtner, 2007). As salient objects to communicate and organize spatial knowledge during navigation, landmarks have attracted the attention of many researchers in the field of spatial cognition and geoinformation science, especially research on landmark‐based pedestrian navigation (Delikostidis, van Elzakker, & Kraak, 2016; Fang, Yang, Guan, Feng, & Jiang, 2020; Filomena & Verstegen, 2021). Landmark visualization is an important factor affecting the efficiency of guidance in landmark‐based pedestrian navigation, but current research is limited in considering the personalized preferences of pedestrians.…”