ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume V-4-2022
https://doi.org/10.5194/isprs-annals-V-4-2022-173-2022
https://doi.org/10.5194/isprs-annals-V-4-2022-173-2022
18 May 2022
 | 18 May 2022

MULTI-RESOLUTION REPRESENTATION USING GRAPH DATABASE

Y. Huang and E. Stefanakis

Keywords: Spatial Information, Geographic Information System, Multi-resolution, Multi-representation, Graph Database, Neo4j

Abstract. Multi-resolution representation has always been an important and popular data source for many research and applications, such as navigation, land cover, map generation, media event forecasting, etc. With one spatial object represented by distinct geometries at different resolutions, multi-resolution representation is high in complexity. Most of the current approaches for storing and retrieving multi-resolution representation are either complicated in structure, or time consuming in traversal and query. In addition, supports on direct navigation between different representations are still intricate in most of the paradigms, especially in topological map sets. To address this problem, we propose a novel approach for storing, querying, and extracting multi-resolution representation. The development of this approach is based on Neo4j, a graph database platform that is famous for its powerful query and advanced flexibility. Benefited from the intuitiveness of the proposed database structure, direct navigation between representations of one spatial object, and between groups of representations at adjacent resolutions are both available. On top of this, collaborating with the self-designed web-based interface, queries within the proposed approach truly embraced the concept of keyword search, which lower the barrier between novice users and complicate queries. In all, the proposed system demonstrates the potential of managing multi-resolution representation data through the graph database and could be a time-saver for related processes.