The necessary yet complex evaluation of 3D city models: a semantic approach

Ennafii, Oussama; Mallet, Clément; Bris, Arnaud Le; Lafarge, Florent

doi:10.1109/jurse.2019.8809002

Cited by 2 publications

(9 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our 3D building model evaluation framework consists in detecting errors, for every building, according to a hierarchical taxonomy of 9 atomic errors [11]. These errors correspond to two distinct levels:…”

Section: The Evaluation Frameworkmentioning

confidence: 99%

“…• [11], where one can see that all sources contribute equally, even if geometric features suffice to reach high accuracies.…”

Section: The Evaluation Frameworkmentioning

confidence: 99%

“…F-score metrics per error are selected for the evaluation of the various configurations. These experiments are also compared with the results obtained for each area (training+testing) with various feature sets [11]: 3D model geometry (Geom. ), VHR optical image (Im.…”

Section: Strategymentioning

confidence: 99%

“…Nantes provides the most effective training set for building errors. Secondly, contrarily to the ablation study conducted in [11], imagebased features and, with a lower impact, height attributes are often decisive in this experimental study. The geometry of the 3D model no longer suffices and remote sensing data is mandatory for transfer learning.…”

Section: Transferabilitymentioning

confidence: 99%

“…Surprisingly, it has been barely investigated in the literature [7,8,9,10]. In [11], we proposed a semantic evaluation framework in which building semantics is taken into account through the detection and categorization of modeling errors at the facet level for each 3D building. Our solution is independent from a given urban area, the Level of Details (LOD) of buildings, and the 3D modeling method.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Scalable Evaluation of 3D City Models

Ennafii

Bris

Lafarge

et al. 2019

IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium

Self Cite

View full text Add to dashboard Cite

The generation of 3D building models from Very High Resolution geospatial data is now an automatized procedure. However, urban areas are very complex and practitioners still have to visually assess the correctness of these models and detect reconstruction errors. We proposed an approach for automatically evaluating the quality of 3D building models. It is cast as a supervised classification task based on a hierarchical taxonomy and multimodal handcrafted features (building geometry, optical images, height data). In this paper, we evaluate how the urban area composition impacts prediction transferability and scalability of our framework to unseen scenes. This allows to define minimal feature and training sets for a problem where no benchmark data has been released so far.

show abstract

Section: The Evaluation Frameworkmentioning

confidence: 99%

“…• [11], where one can see that all sources contribute equally, even if geometric features suffice to reach high accuracies.…”

Section: The Evaluation Frameworkmentioning

confidence: 99%

Section: Strategymentioning

confidence: 99%

Section: Transferabilitymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Scalable Evaluation of 3D City Models

Ennafii

Bris

Lafarge

et al. 2019

IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium

Self Cite

View full text Add to dashboard Cite

show abstract

A Learning Approach to Evaluate the Quality of 3D City Models

Ennafii¹,

Bris²,

Lafarge³

et al. 2019

photogramm eng remote sensing

Self Cite

View full text Add to dashboard Cite

The automatic generation of three-dimensional (3D) building models from geospatial data is now a standard procedure. An abundance of literature covers the last two decades, and several solutions are now available. However, urban areas are very complex environments. Inevitably, practitioners still have to visually assess, at a city-scale, the correctness of these models and detect frequent reconstruction errors. Such a process relies on experts and is highly time-consuming, with approximately two hours/km 2 per expert. This work proposes an approach for automatically evaluating the quality of 3D building models. Potential errors are compiled in a novel hierarchical and versatile taxonomy. This allows, for the first time, to disentangle fidelity and modeling errors, whatever the level of details of the modeled buildings. The quality of models is predicted using the geometric properties of buildings and, when available, Very High Resolution images and Digital Surface Models. A baseline of handcrafted, yet generic, features is fed into a Random Forest classifier. Both multiclass and multilabel cases are considered: due to the interdependence between classes of errors, it is possible to retrieve all errors at the same time while simply predicting correct and erroneous buildings. The proposed framework was tested on three distinct urban areas in France with more than 3000 buildings. 80%–99% F-score values are attained for the most frequent errors. For scalability purposes, the impact of the urban area composition on the error prediction was also studied, in terms of transferability, generalization, and representativeness of the classifiers. It showed the necessity of multimodal remote sensing data and mixing training samples from various cities to ensure a stability of the detection ratios, even with very limited training set sizes.

show abstract

The necessary yet complex evaluation of 3D city models: a semantic approach

Cited by 2 publications

References 5 publications

Scalable Evaluation of 3D City Models

Scalable Evaluation of 3D City Models

A Learning Approach to Evaluate the Quality of 3D City Models

Contact Info

Product

Resources

About