Premium
A data model for organizing relative semantics as images to support pedestrian navigation computations
Author(s) -
Fang Zhixiang,
Yang Fan,
Guan Fangli,
Feng Mingxiang,
Jiang Yuxin
Publication year - 2020
Publication title -
transactions in gis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.721
H-Index - 63
eISSN - 1467-9671
pISSN - 1361-1682
DOI - 10.1111/tgis.12669
Subject(s) - computer science , semantics (computer science) , pedestrian , computation , data mining , reference frame , information retrieval , frame (networking) , algorithm , transport engineering , engineering , programming language , telecommunications
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.