Quantitative expressions of spatial similarity between road networks in multiscale map spaces

ABSTRACT

Spatial similarity plays a critical role in the perception and cognition in capturing information from maps; it can be used as a constraint to automate map generalization. Although measuring similarities seems natural to humans, it can be challenging to quantify them. This is especially true when it comes to calculating spatial similarity degrees between groups of spatial objects at varying scales and quantitatively expressing the relations between spatial similarity and change of map scale in multiscale map spaces. Taking road networks as an example, this paper proposes an approach to measuring spatial similarity between a road network at a large scale and its generalized counterpart at a smaller scale. By fitting a power function to three typical types of road networks, this paper provides a formula for expressing the change in spatial similarity as the map scale changes. The proposed quantitative method lays a foundation for using spatial similarity as a constraint during road network generalization.

RÉSUMÉ

La similarité spatiale joue un rôle essentiel dans la perception et la cognition pour acquérir de l'information à partir de cartes. Elle peut être utilisée comme une contrainte pour automatiser la généralisation cartographique. Bien que mesurer des similarités semble naturel pour les humains, leur quantification peut être un défi. C'est particulièrement vrai quand il s'agit de calculer des degrés de similarité spatiale entre groupes d'objets spatiaux à différentes échelles et d'exprimer quantitativement les relations entre la similarité spatiale et le changement d'échelle cartographique dans un espace de cartes multi-échelles. En prenant l'exemple des réseaux routiers, cet article propose une approche pour mesure la similarité spatiale entre un réseau routier à grande échelle et son homologue généralisé à une plus petite échelle. En ajustant une fonction de puissance pour trois classes standard de réseau de routes, cet article propose une formule pour exprimer le changement de similarité spatiale lorsque l'échelle de la carte change. La méthode quantitative proposée repose sur l'utilisation de la similarité spatiale comme une contrainte pendant la généralisation de réseaux routiers.

KEYWORDS:

• Spatial similarity
• road networks
• constraint
• map generalization
• multiscale map spaces

Acknowledgments

The authors sincerely thank the editor and the anonymous reviewers for their insightful comments and valuable suggestions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [grant number 41930101, 42161066].

Notes on contributors

Haowen Yan

Haowen Yan is a professor of geographic information science with Lanzhou Jiaotong University. He is the editor-in-chief of Journal of Geovisualization and Spatial Analysis (since 2016). His research interests are in map generalization, spatial analysis and geovisualization.

Weifang Yang

Weifang Yang is a professor at the Faculty of Geomatics, Lanzhou Jiaotong University, specializing in surveying engineering. Her research focuses on surveying instrument metrology, surveying data analytics and GNSS meteorology.

Xiaomin Lu

Xiaomin Lu is an associate professor of geographic information science and associate director of GIS department at Lanzhou Jiaotong University. Her research interests are map generalization and spatial relations.

Pengbo Li

Pengbo Li is currently a Ph.D. candidate at the Faculty of Geomatics, Lanzhou Jiaotong University, Lanzhou, China. His research interests include map generalization and machine learning.