OpenStreetMap – building data completeness visualization in terms of “Fitness for purpose”

• Published in: Advances in Geodesy and Geoinformation Vol. 72; no. No 1; p. 35
• Main Authors: Borkowska, Sylwia, Bielecka, Elzbieta, Pokonieczny, Krzysztof
• Format: Journal Article
• Language: English
• Published: Polish Academy of Sciences 26.06.2023
• Subjects: bivariate choropleth map; data completeness; data quality; osm; vgi
• ISSN: 2720-7242; 2720-7242
• DOI: 10.24425/agg.2022.141922

The purpose of this article was to provide the user with information about the number of buildings in the analyzed OpenStreetMap (OSM) dataset in the form of data completeness indicators, namely the standard OSM building areal completeness index (C Index), the numerical completeness index (COUNT Index) and OSM building location accuracy index (TP Index). The official Polish vector database BDOT10k (Database of Topographic Objects) was designated as the reference dataset. Analyses were carried out for Piaseczno County in Poland, differentiated by land cover structure and urbanization level. The results were presented in the form of a bivariate choropleth map with an individually selected class interval suitable for the statistical distribution of the analyzed data. The results confirm that the completeness of OSM buildings close to 100% was obtained mainly in built-up areas. Areas with a commission of OSM buildings were distinguished in terms of area and number of buildings. Lower values of completeness rates were observed in less urbanized areas. The developed methodology for assessing the quality of OSM building data and visualizing the quality results to assist the user in selecting a dataset is universal and can be applied to any OSM polygon features, as well as for peer review of other spatial datasets of comparable thematic scope and detail.