An improved LOD specification for 3D building models

https://doi.org/10.1016/j.compenvurbsys.2016.04.005Get rights and content

Highlights

  • CityGML LODs are an industry standard for conveying the grade of 3D city models.

  • The 5 LODs are not defined precisely, and they are not sufficient for this purpose.

  • We present a refined series of 16 LODs that overcomes these issues.

Abstract

The level of detail (LOD) concept of the OGC standard CityGML 2.0 is intended to differentiate multi-scale representations of semantic 3D city models. The concept is in practice principally used to indicate the geometric detail of a model, primarily of buildings. Despite the popularity and the general acceptance of this categorisation, we argue in this paper that from a geometric point of view the five LODs are insufficient and that their specification is ambiguous.

We solve these shortcomings with a better definition of LODs and their refinement. Hereby we present a refined set of 16 LODs focused on the grade of the exterior geometry of buildings, which provide a stricter specification and allow less modelling freedom. This series is a result of an exhaustive research into currently available 3D city models, production workflows, and capabilities of acquisition techniques. Our specification also includes two hybrid models that reflect common acquisition practices. The new LODs are in line with the LODs of CityGML 2.0, and are intended to supplement, rather than replace the geometric part of the current specification. While in our paper we focus on the geometric aspect of the models, our specification is compatible with different levels of semantic granularity. Furthermore, the improved LODs can be considered format-agnostic.

Among other benefits, the refined specification could be useful for companies for a better definition of their product portfolios, and for researchers to specify data requirements when presenting use cases of 3D city models. We support our refined LODs with experiments, proving their uniqueness by showing that each yields a different result in a 3D spatial operation.

Introduction

The level of detail (LOD) of a 3D city model is one of its most important characteristics. It denotes the adherence of the model to its real-world counterpart, and it has implications on its usability (Biljecki, Ledoux, Stoter, & Zhao, 2014b).

The CityGML 2.0 standard from the Open Geospatial Consortium (2012) defines five LODs. The concept is intended for several thematic classes of objects but it is primarily focused on buildings, and the five described instances increase in their geometric and semantic complexity (Fig. 1). LOD0 is a representation of footprints and optionally roof edge polygons marking the transition from 2D to 3D GIS. LOD1 is a coarse prismatic model usually obtained by extruding an LOD0 model. LOD2 is a model with a simplified roof shape, and where the object's parts can be modelled in multiple semantic classes (e.g. roof, wall). LOD3 is an architecturally detailed model with windows and doors, being considerably more complex than its preceding counterpart. LOD4 completes an LOD3 by including indoor features (Kolbe, 2009). This taxonomy has been developed in the German Special Interest Group 3D (SIG 3D) initiative (Albert, Bachmann, & Hellmeier, 2003), and has been further described in Gröger and Plümer (2012). The five LODs have become widely adopted by the stakeholders in the 3D GIS industry and they now also describe the grade and the design quality of a 3D city model, especially its geometric aspect (i.e. “how much detail should be acquired?”). They have gained importance also in the computer graphics (Verdie et al., 2015, Musialski et al., 2013), and BIM (Tolmer, Castaing, Diab, & Morand, 2013) communities when dealing with 3D building models.

The LOD concept of CityGML is primarily intended to differentiate the grade of data resulting from different production workflows, and they are driven by semantics as much as geometry. In the industry and research community they were accepted from the outlook on geometric richness, which was partly caused by the lack of applications that require semantics. For instance, we have observed that while the LOD2 from the point of view of CityGML developers represents a model with differentiated semantic surfaces, practitioners primarily refer to models with roof shapes, even when not dealing with data that is semantically structured.

While the five LODs generally provide a categorisation of the overall level of abstraction, content, value, and usability of 3D city models, this classification has several drawbacks and shortcomings as we show in Section 2. Since the specification is crucial among practitioners and researchers for conveying the grade of a 3D city model and its adherence to the real-world, in this paper we present a refined specification to solve such problems. It should be noticed that the topic of refining and improving the current specification of the LODs is currently under consideration in the CityGML community for version 3.0 (Machl, 2013, Löwner and Gröger, 2016), and we hope that our proposal will help the discussions. However, our work is intended to be independent of any particular 3D format, and applicable to any format that can be used to store 3D building models, including ones such as COLLADA and OBJ.

In Fig. 2 we give an example of the shortcomings of the current concept, from the point of view of the geometric detail. The figure illustrates two LOD2 models: the model on the left has been acquired with two acquisition techniques, the walls are at their actual location and the roof overhangs are explicitly present. The representation in the middle has been acquired with one technique (aerial photogrammetry) where the walls are derived as projections from the roof outline (the third model will be introduced in another example in the following section). This example illustrates how the CityGML LOD concept is ambiguous and that it falls short in defining the complexity of the models: the two models are of the same LOD (LOD2) according to CityGML while the first one is more laborious to acquire and it may bring better results in a spatial analysis (e.g. more accurate volume; see Biljecki, Ledoux, Stoter, & Vosselman, 2016). Hence, practitioners would not consider them to be of equal value and usability. For these reasons we argue in this paper that they should be considered as different LODs, and our specification differentiates such cases.

This ambiguity is most evident in the production of the models. For instance, in 3D generalisation where researchers produce multiple geometric variants of LODs and discuss the ambiguity, among others see Guercke, Götzelmann, Brenner, and Sester (2011), Fan and Meng (2012), Stoter et al. (2011), Noskov and Doytsher (2014), and Deng et al. (2016).

Solving the ambiguity is also important considering: (1) the increasing number of acquisition techniques (e.g. the recently investigated being drones (Nex & Remondino, 2013), radar (Zhu & Shahzad, 2014), handheld devices (Rosser et al., 2015, Sirmacek and Lindenbergh, 2014), procedural modelling (Wonka et al., 2003, Müller et al., 2006, Kelly and Wonka, 2011, Müller Arisona et al., 2013, Tsiliakou et al., 2014, Smelik et al., 2014), conversion from BIM and computer graphics models (Donkers et al., 2015, Kumar et al., 2016), and generation from 2D drawings (Gimenez, Hippolyte, Robert, Suard, & Zreik, 2015)); (2) the number of data producers and national mapping agencies requesting 3D data is increasing (Stoter et al., 2015), and without a finer specification data producers and users may resort to creating their own specifications (e.g. see the series from Blom, 2011), which might increase the ambiguity; (3) the increase in quantity of data sets with non-homogenous LODs (Fan et al., 2014, Touya and Reimer, 2015, Arroyo Ohori et al., 2015a); and (4) use cases have different requirements when it comes to the complexity and quality of the data. Furthermore, the number of 3D use cases is rapidly increasing (Biljecki, Stoter, Ledoux, Zlatanova, & Çöltekin, 2015b), for instance — solar potential estimation (Freitas, Catita, Redweik, & Brito, 2015), studying the thermal characteristics of the outdoor space (Maragkogiannis, Kolokotsa, Maravelakis, & Konstantaras, 2014), firefighting simulations (Chen, Wu, Shen, & Chou, 2014), and advances in multi-scale navigation (Hildebrandt & Timm, 2014). Each of these use cases may have different requirements when it comes to the LOD of the models.

In this paper we improve the geometric aspect of the LOD specification of 3D building models. We provide an extended and more informative series of 16 LODs that are compatible with the existing CityGML LODs. The refined taxonomy is a result of a research into currently available 3D city models and an investigation of the acquisition workflows. We review related work on this topic (Section 3), and for each LOD we give requirements and show an example (Section 4).

We have generated a sample data set in 16 LODs and run them through a few GIS operations to show that each LOD is unique from a geometric point of view and may bring different results in a spatial analysis (Section 5).

In this paper we focus on the exterior of buildings (i.e. their exterior shell in LOD0–3). The refinement of the indoor and semantics aspect of the specification can be considered as orthogonal topics to this one. These topics are being tackled by other researchers who decompose it into different levels of abstraction and integrate them into expanded LOD1, LOD2 and LOD3 models (for examples see the work of Boeters, Arroyo Ohori, Biljecki, and Zlatanova (2015) and Löwner, Benner, Gröger, and Häfele (2013)). While the semantic LOD and indoor LOD are out of scope of our paper, present work on these topics is compatible with our work because such specification can be supplemented to ours. For instance, each of the newly refined LODs can be assigned a semantic LOD depending on the achieved spatio-semantic coherence.

Section snippets

Shortcomings of the current concept and difficulties with designing a specification

The LOD concept used in CityGML 2.0 has been borrowed from computer graphics in which multiple representations of polygon meshes are differentiated by their number of faces, and their simplification is performed by algorithms that reduce the number of faces while attempting to retain visual fidelity (Luebke et al., 2003, Clark, 1976). While the early purpose of the LOD in 3D GIS was to improve visualisation performance (e.g. see Coors & Flick, 1998), visualisation is now only one of the

Related work

The general LOD notion was examined in our earlier work (Biljecki et al., 2014b). The concept is decomposed into six metrics: list of features, their geometric complexity, dimensionality, appearance, spatio-semantic coherence, and attributes. We take into account the first three metrics when defining the geometric aspect of the LOD.

Stoter et al. (2011) recognise that CityGML lacks precise LOD definitions and allows ambiguity, and in a later research, Stoter, Vosselman, Dahmen, Oude Elberink,

Refined levels of detail for buildings

We provide a series that contains 16 LODs (4 refined LODs for each of the LOD0–3), which are shaped after a literature review and inventory of presently available models by finding their main relevant similarities and mutual aspects. A visual example of the refined LODs is shown in Fig. 3. We believe that these LODs allow for less ambiguity, and they aid practitioners to standardise their data with an improved definition of the complexity of the models. As mentioned before, this work is not

Proving the specification with geometric experiments

For the experiments, we have generated a CityGML data set of 100 buildings in the 16 LODs we propose. The data has been generated procedurally with a CityGML engine developed by Biljecki, Ledoux, and Stoter (2014a), and it has been converted to the OBJ file format to broaden their usability. The conversion was done with the tool CityGML2OBJs (Biljecki & Arroyo Ohori, 2015), and in the process the models were triangulated with Triangle (Shewchuk, 1996).

We have computed the following quantitative

Conclusions and future work

The current LOD categorisation of CityGML has two shortcomings:

  • 1.

    lack of a precise specification of each LOD; and that

  • 2.

    the current five LODs are too generic and therefore they are not always capable to separate one LOD from the other (i.e. two significantly different levels of abstraction may still be considered as the same LOD as per the current specification).

The refined LODs that we have introduced are a result of a literature review, analysing acquisition workflows, and examining publicly

Acknowledgements

We gratefully acknowledge the comments of the members of the EuroSDR 3D Special Interest Group, members of the OGC CityGML Standard Working Group, members of the OGC CityGML Quality Interoperability Experiment, and those of the anonymous reviewers. We appreciate the input given by companies specialised in the acquisition of 3D city models.

This research is supported by the Dutch Technology Foundation STW, which is part of The Netherlands Organisation for Scientific Research (NWO), and which is

References (208)

  • V. Coors

    3D-GIS in networking environments

    Computers, Environment and Urban Systems

    (2003)
  • M. De Berg et al.

    On levels of detail in terrains

    Graphical Models and Image Processing

    (1998)
  • Y. Deng et al.

    Mapping between BIM and 3D GIS in different levels of detail using schema mediation and instance comparison

    Automation in Construction

    (2016)
  • U. Eicker et al.

    Assessing passive and active solar energy resources in cities using 3D city models

    Energy Procedia

    (2014)
  • K. Fath et al.

    A method for predicting the economic potential of (building-integrated) photovoltaics in urban areas based on hourly radiance simulations

    Solar Energy

    (2015)
  • A. Forberg

    Generalization of 3D building data based on a scale-space approach

    ISPRS Journal of Photogrammetry and Remote Sensing

    (2007)
  • S. Freitas et al.

    Modelling solar potential in the urban environment: State-of-the-art review

    Renewable and Sustainable Energy Reviews

    (2015)
  • L. Gimenez et al.

    Review: Reconstruction of 3D building information models from 2D scanned plans

    Journal of Building Engineering

    (2015)
  • T. Glander et al.

    Abstract representations for interactive visualization of virtual 3D city models

    Computers, Environment and Urban Systems

    (2009)
  • G. Gröger et al.

    CityGML — Interoperable semantic 3D city models

    ISPRS Journal of Photogrammetry and Remote Sensing

    (2012)
  • R. Guercke et al.

    Aggregation of LoD 1 building models as an optimization problem

    ISPRS Journal of Photogrammetry and Remote Sensing

    (2011)
  • N. Haala et al.

    An update on automatic 3D building reconstruction

    ISPRS Journal of Photogrammetry and Remote Sensing

    (2010)
  • A. Henn et al.

    Model driven reconstruction of roofs from sparse LIDAR point clouds

    ISPRS Journal of Photogrammetry and Remote Sensing

    (2013)
  • AdV

    Produktstandard für 3D-Gebäudemodelle

  • AdV

    Modellierungsbeispiele für 3D-Gebäudemodelle

  • G. Agugiaro

    From sub-optimal datasets to a CityGML-compliant 3D city model: Experiences from Trento, Italy

    The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences

    (2014)
  • F.C. Ahmed et al.

    Using three-dimensional volumetric analysis in everyday urban planning processes

    Applied Spatial Analysis and Policy

    (2015)
  • R. Akmalia et al.

    TLS for generating multi-LOD of 3D building model

  • N. Alam et al.

    Detecting shadow for direct radiation using CityGML models for photovoltaic potentiality analysis

  • J. Albert et al.

    Zielgruppen und Anwendungen für Digitale Stadtmodelle und Digitale Geländemodell

  • J.H. Amorim et al.

    Detailed modelling of the wind comfort in a city avenue at the pedestrian level

  • K.H. Anders

    Level of detail generation of 3D building groups by aggregation and typification

  • C. Andujar et al.

    Visualization of large-scale urban models through multi-level relief impostors

    Computer Graphics Forum

    (2010)
  • K. Aringer et al.

    Bavarian 3D building model and update concept based on LiDAR, image matching and cadastre information

  • K. Arroyo Ohori et al.

    A dimension-independent extrusion algorithm using generalised maps

    International Journal of Geographical Information Science

    (2015)
  • K. Arroyo Ohori et al.

    Modeling a 3D city model and its levels of detail as a true 4D model

    ISPRS International Journal of Geo-Information

    (2015)
  • J.M. Bahu et al.

    Towards a 3D spatial urban energy modelling approach

  • S.U. Baig et al.

    Generalization of buildings within the framework of CityGML

    Geo-spatial Information Science

    (2013)
  • M. Batty et al.

    Visualizing the city: communicating urban design to planners and decision-makers

  • S. Becker

    Towards complete LOD3 models — Automatic interpretation of building structures

  • N. Ben Fekih Fradj et al.

    Abschätzung des nutzbaren Dachflächenanteils für Solarenergie mit CityGML-Gebäudemodellen und Luftbildern

  • J. Benner et al.

    Enhanced LOD concepts for virtual 3D city models

  • Benner, J., Geiger, A., Häfele, K.H., 2010. Concept for building licensing based on standardized 3d geo information....
  • L.A.H.M. van Berlo et al.

    Creating the Dutch national BIM levels of development

  • G. Besuievsky et al.

    A configurable LoD for procedural urban models intended for daylight simulation

  • Z. Biljecki

    Concept and implementation of Croatian Topographic Information System—CROTIS

    (2007)
  • F. Biljecki et al.

    Automatic semantic-preserving conversion between OBJ and CityGML

  • F. Biljecki et al.

    Propagation of positional error in 3D GIS: estimation of the solar irradiation of building roofs

    International Journal of Geographical Information Science

    (2015)
  • F. Biljecki et al.

    Error propagation in the computation of volumes in 3D city models with the Monte Carlo method

  • F. Biljecki et al.

    Applications of 3D city models: State of the art review

    ISPRS International Journal of Geo-Information

    (2015)
  • Cited by (328)

    View all citing articles on Scopus
    View full text