Point-feature label placement with maximum entropy principle
Zhiwei Wu Zhilin Li Tian Lan a Faculty of Geosciences and Engineering,Southwest Jiaotong University,Chengdu,Chinab Moganshan Geospatial Information Lab,Deqing,ChinaZhiwei Wu is a Ph.D. candidate at Southwest Jiaotong University. His research interests include AI cartography,automated digital cartography,and Spatial complexity measure. He contributed to the conceptualization,data processing,experiment analysis,visualization,and writing (original draft,review & editing).Zhilin Li is a professor at Southwest Jiaotong University. His research interests include digital cartography,scale-driven spatial data modelling and analysis,and feature extraction from remote sensing images. He contributed to the conceptualization,methodology,writing (review & editing),resources,and funding acquisition.Tian Lan is an assistant professor at Southwest Jiaotong University. His research interests include AI cartography,automated digital cartography,and spatial complex network analysis. He contributed to the conceptualization,writing (review & editing),and funding acquisition.
DOI: https://doi.org/10.1080/13658816.2024.2422502
2024-11-05
International Journal of Geographical Information Science
Abstract:A map becomes readable and translatable only after the use of labels. High-quality label placement (i.e., labelling) is a combinatorial optimization problem, where one or more objective functions are required. However, such objective functions are still not well achieved in existing methods with commonly used labelling rules (e.g., "avoidance of overlapping labels" and "placement at priority positions"). In this study, we define a new objective function using the maximum entropy principle and employ a modified genetic algorithm in the optimization process. Comparative experiments have been conducted with 29 existing labelling methods in a dataset of 1000 points. Experimental results show that the average percentage of non-overlapping labels by this new method reaches 94.73% (i.e., being ranked second in all 30 methods) and has no statistically significant difference from the best one (i.e., 94.87%). By further analysis, it is found that this new method not only can place more labels in priority positions than the method being ranked first but also has the potential to place labels for line and area features. Such results indicate that the maximum entropy principle may pave the way for further research in cartography.
geography, physical,computer science, information systems,information science & library science