Generating 3D city models without elevation data. (July 2017)

• Record Type: Journal Article
• Authors: Biljecki, Filip; Ledoux, Hugo; Stoter, Jantien
• Is Part Of: Computers, environment and urban systems. Volume 64(2017)
• Page Start: 1
• Page End: 18
• Publication Date: 2017-07
• Subjects: 3D city models -- GIS -- Building height -- Lidar -- Urban models -- Urban morphology -- Random forest -- CityGML -- LOD1; City planning -- Data processing -- Periodicals; Regional planning -- Data processing -- Periodicals; 303.4834
• DOI: 10.1016/j.compenvurbsys.2017.01.001
• Languages: English
• ISSNs: 0198-9715

Abstract

Abstract: Elevation datasets (e.g. point clouds) are an essential but often unavailable ingredient for the construction of 3D city models. We investigate in this paper to what extent can 3D city models be generated solely from 2D data without elevation measurements. We show that it is possible to predict the height of buildings from 2D data (their footprints and attributes available in volunteered geoinformation and cadastre), and then extrude their footprints to obtain 3D models suitable for a multitude of applications. The predictions have been carried out with machine learning techniques (random forests) using 10 different attributes and their combinations, which mirror different scenarios of completeness of real-world data. Some of the scenarios resulted in surprisingly good performance (given the circumstances): we have achieved a mean absolute error of 0.8m in the inferred heights, which satisfies the accuracy recommendations of CityGML for LOD1 models and the needs of several GIS analyses. We show that our method can be used in practice to generate 3D city models where there are no elevation data, and to supplement existing datasets with 3D models of newly constructed buildings to facilitate rapid update and maintenance of data. Graphical Abstract: Highlights: Lack of elevation data hinders the construction of 3D city models. We infer heights of buildings solely from 2D footprints and attributes. LOD1 models are generated by extruding footprints to the predicted height. We achieve sub-meter accuracy in the predicted heights. The resulting 3D models satisfy the CityGML standard quality recommendations and those of several spatial analyses.