# Large-Scale Accurate Reconstruction of Buildings Employing Point Clouds Generated from UAV Imagery

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

- New data: utilizing high-density photogrammetric point clouds as the only source of data. According to the literature review, most of the papers exploit photogrammetric point clouds as an additional source of data. The capabilities of this type of point cloud in building reconstruction have not been much appreciated.
- Improving segmentation method: point-based analyses are performed to improve the segmentation method in sizing up segments.
- New constrained planar modelling: the modelling method is developed to create models of facades, roofs and the grounds by imposing the constraints on local normal vectors. The plane-based generation of footprints is a novel usage of points on walls and grounds in building reconstruction.
- New evaluation method: this method examines segmentation in the viewpoint of generation of planar segments and proposes vertex-based criteria to explore over/under segmentation and refine the reconstructed footprint.

## 2. Overview of the Method and Background

#### 2.1. Segmentation

#### 2.2. Constrained Least Squares Modelling

## 3. Materials and Methods

^{2}and 1240 m

^{2}. The average density of the point clouds on roofs is 30 points/dm

^{2}. Furthermore, the average density on walls is 20 points/dm

^{2}. This amount for grounds is 24 points/dm

^{2}.These data are intricate and bear many details, clutter, noises and holes.

#### 3.1. Two-Level Segmentation

#### Improvement in the Efficient RANSAC

#### 3.2. Planar Modelling

## 4. Results

## 5. Discussion

## 6. Summary and Conclusions

## Author Contributions

## Acknowledgments

## Conflicts of Interest

## Appendix A

Properties | Data |
---|---|

Focal length | 35 mm |

Resolution | 7360 × 4912 |

Pixel size | 4.89 × 4.89 µm |

Fx | 7429.05 pixel |

Fy | 7429.05 pixel |

k1 | 0.0557213 |

k2 | −0.219064 |

k3 | −0.0120905 |

k4 | 0 |

skew | 0 |

cx | 3650.54 |

cy | 2463.08 |

P1 | −0.000600474 |

P2 | 0.000548429 |

## Appendix B

**Figure A1.**Approximation of the step roof’s number of points. (

**a**,

**b**) Recognizing the edges of the roof as the borders of half-planes via histogram of $\raisebox{1ex}{${\lambda}_{2}$}\!\left/ \!\raisebox{-1ex}{${\lambda}_{3}$}\right.$, and (

**c**) generation of the main edges of the roof. (Units of axes in (

**b**,

**c**) are m).

## Appendix C

Parameter | Values |
---|---|

Maximum distance to the primitive (m) | 0.01–0.10 |

Maximum normal deviation (degree) | 1–5 |

Sampling resolution (m) | 0.1 |

Overlooking probability (%) | 95 |

Minimum points per primitive | ${n}_{{p}_{i}}$ |

## References

- Haala, N.; Kada, M. An update on automatic 3D building reconstruction. J. Photogramm. Remote Sens.
**2010**, 65, 570–580. [Google Scholar] [CrossRef] - Gröger, G.; Kolbe, T.H.; Nagel, C.; Häfele, K.-H. OGC City Geography Markup Language (CityGML) Encoding Standard. 2012. Available online: https://portal.opengeospatial.org/files/?artifact_id=47842 (accessed on 1 January 2018).
- Donkers, S. Automatic Generation of CityGML LoD3 Building Models from IFC Models. Master’s Thesis, TU Delft, Delft, The Netherlands, 2013. [Google Scholar]
- Löwner, M.; Benner, J.; Gröger, G.; Häfele, K. New Concepts for Structuring 3D City Models—An Extended Level of Detail Concept for CityGML Buildings. In Computational Science and Its Applications—ICCSA 2013; Lecture Notes in Computer Science; Springer: Berlin/Heidelberg, Germany, 2013; Volume 7973. [Google Scholar]
- Tarsha-Kurdi, F.; Landes, T.; Grussenmeyer, P.; Koehl, M. Model-driven and data-driven approaches using LIDAR data: Analysis and comparison. Int. Arch. Photogramm. Remote Sens.
**2007**, 36, 87–92. [Google Scholar] - Elberink, S.O.; Vosselman, G. Building reconstruction by target based graph matching on incomplete laser data: Analysis and limitations. Sensors
**2009**, 9, 6101–6118. [Google Scholar] [CrossRef] [PubMed] - Wichmann, A.; Jung, J.; Sohn, G.; Kada, M.; Ehlers, M. Integration of building knowledge into binary space partitioning for the reconstruction of regularized building models. ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci.
**2015**, II-3/W5, 541–548. [Google Scholar] [CrossRef] - Satari, M.; Samadzadegan, F.; Azizi, A.; Maas, H.-G. A multi-resolution hybrid approach for building model reconstruction from LIDAR data. Photogramm. Rec.
**2012**, 27, 330–359. [Google Scholar] [CrossRef] - Zolanvari, S.M.I.; Laefer, D.F.; Natanzi, A.S. Three-dimensional building façade segmentation and opening area detection from point clouds. ISPRS J. Photogramm. Remote Sens.
**2018**. [Google Scholar] [CrossRef] - Pahlavani, P.; Amini Amirkolaee, H.; Bigdeli, B. 3D reconstruction of buildings from LiDAR data considering various types of roof structures. Int. J. Remote Sens.
**2017**, 38, 1451–1482. [Google Scholar] [CrossRef] - Jung, J.; Jwa, Y.; Sohn, G. Implicit regularization for reconstructing 3D building rooftop models using airborne LiDAR data. Sensors
**2017**, 17, 621. [Google Scholar] [CrossRef] [PubMed] - Huang, H.; Brenner, C.; Sester, M. A generative statistical approach to automatic 3D building roof reconstruction from laser scanning data. ISPRS J. Photogramm. Remote Sens.
**2013**, 79, 29–43. [Google Scholar] [CrossRef] - Wang, F.; Xi, X.; Wang, C.; Xiao, Y.; Wan, Y. Boundary regularization and building reconstruction based on terrestrial laser scanning data. In Proceedings of the 2013 IEEE International Geoscience and Remote Sensing Symposium—IGARSS, Melbourne, Australia, 21–26 July 2013; pp. 1509–1512. [Google Scholar]
- Hammoudi, K. Contributions to the 3D City Modeling: 3D Polyhedral Building Model Reconstruction from Aerial Images and 3D Facade Modeling from Terrestrial 3D Point Cloud and Images; Université Paris-Est: Champs-sur-Marne, France, 2011. [Google Scholar]
- Rupnik, E.; Pierrot-Deseilligny, M.; Delorme, A. 3D reconstruction from multi-view VHR-satellite images in MicMac. ISPRS J. Photogramm. Remote Sens.
**2018**, 139, 201–211. [Google Scholar] [CrossRef] - Haala, N.; Rothermel, M.; Cavegn, S. Extracting 3D urban models from oblique aerial images. In Proceedings of the 2015 Joint Urban Remote Sensing Event (JURSE), IEEE, Lausanne, Switzerland, 30 March–1 April 2015; pp. 1–4. [Google Scholar]
- Xiao, J.; Gerke, M.; Vosselman, G. Building extraction from oblique airborne imagery based on robust façade detection. ISPRS J. Photogramm. Remote Sens.
**2012**, 68, 56–68. [Google Scholar] [CrossRef] - Dahlke, D.; Linkiewicz, M.; Meissner, H. True 3D building reconstruction: Façade, roof and overhang modelling from oblique and vertical aerial imagery. Int. J. Image Data Fusion
**2015**, 6, 314–329. [Google Scholar] [CrossRef] - Vetrivel, A.; Gerke, M.; Kerle, N.; Vosselman, G. Segmentation of UAV-based images incorporating 3D point cloud information. In Proceedings of the Photogrammetric Image Analysis, PIA15 + High-Resolution Earth Imaging for Geospatial Information, HRIGI15: Joint ISPRS Conference, Munich, Germany, 25–27 March 2015; Volume XL-3/W2, pp. 261–268. [Google Scholar]
- Xu, Z.; Wu, L.; Zhang, Z. Use of active learning for earthquake damage mapping from UAV photogrammetric point clouds. Int. J. Remote Sens.
**2017**, 8, 1–28. [Google Scholar] [CrossRef] - Henn, A.; Gröger, G.; Stroh, V.; Plümer, L. Model driven reconstruction of roofs from sparse LIDAR point clouds. ISPRS J. Photogramm. Remote Sens.
**2013**, 76, 17–29. [Google Scholar] [CrossRef] - Fu, K.; Zhang, Y.; Sun, X.; Diao, W.; Wu, B.; Wang, H. Automatic building reconstruction from high resolution InSAR data using stochastic geometrical model. In Proceedings of the 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Beijing, China, 10–15 July 2016; pp. 1579–1582. [Google Scholar]
- Salehi, A.; Mohammadzadeh, A. Building Roof Reconstruction Based on Residue Anomaly Analysis and Shape Descriptors from Lidar and Optical Data. Photogramm. Eng. Remote Sens.
**2017**, 83, 281–291. [Google Scholar] [CrossRef] - Yang, L.; Sheng, Y.; Wang, B. 3D reconstruction of building facade with fused data of terrestrial LiDAR data and optical image. Opt. Int. J. Light Electron Opt.
**2016**, 127, 2165–2168. [Google Scholar] [CrossRef] - Pu, S.; Vosselman, G. Building facade reconstruction by fusing terrestrial laser points and images. Sensors
**2009**, 9, 4525–4542. [Google Scholar] [CrossRef] [PubMed] - Ameri, B.; Fritsch, D. Automatic 3D building reconstruction using plane-roof structures. In Proceedings of the American Society for Photogrammetry and Remote Sensing, Washington, DC, USA, 22–26 May 2000. [Google Scholar]
- Malihi, S.; Valadan Zoej, M.J.; Hahn, M.; Mokhtarzade, M.; Arefi, H. 3D building reconstruction using dense photogrammetric point cloud. ISPRS Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci.
**2016**, XLI-B3, 71–74. [Google Scholar] [CrossRef] - Haala, N.; Cavegn, S. High density aerial image matching: State-of-the-art and future prospects. ISPRS Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci.
**2016**, XLI-B4, 625–630. [Google Scholar] [CrossRef] - Ressl, C.; Brockmann, H.; Mandlburger, G.; Pfeifer, N. Dense Image Matching vs. Airborne Laser Scanning—Comparison of two methods for deriving terrain models. Photogramm. Fernerkund. Geoinf.
**2016**, 2016, 57–73. [Google Scholar] [CrossRef] - Cavegn, S.; Haala, N.; Nebiker, S.; Rothermel, M.; Tutzauer, P. Benchmarking high density image matching for oblique airborne imagery. ISPRS Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci.
**2014**, XL-3, 45–52. [Google Scholar] [CrossRef] - Xiong, B.; Oude Elberink, S.; Vosselman, G. Building modeling from noisy photogrammetric point clouds. ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci.
**2014**, II-3, 197–204. [Google Scholar] [CrossRef] - Tutzauer, P.; Haala, N. Façade reconstruction using geometric and radiometric point cloud information. ISPRS Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci.
**2015**, XL-3/W2, 247–252. [Google Scholar] [CrossRef] - Nex, F.; Remondino, F. UAV for 3D mapping applications: A review. Appl. Geomat.
**2014**, 6, 1–15. [Google Scholar] [CrossRef] - Colomina, I.; Molina, P. Unmanned aerial systems for photogrammetry and remote sensing: A review. ISPRS J. Photogramm. Remote Sens.
**2014**, 92, 79–97. [Google Scholar] [CrossRef] - Kolbe, T.H.; Gröger, G.; Plümer, L. CityGML: Interoperable access to 3D city models. In Geo-information for Disaster Management; Springer: Berlin/Heidelberg, Germany, 2005; pp. 883–899. [Google Scholar]
- Awrangjeb, M. Using point cloud data to identify, trace, and regularize the outlines of buildings. Int. J. Remote Sens.
**2016**, 37, 551–579. [Google Scholar] [CrossRef] [Green Version] - Dai, Y.; Gong, J.; Li, Y.; Feng, Q. Building segmentation and outline extraction from UAV image-derived point clouds by a line growing algorithm. Int. J. Digit. Earth
**2017**, 10, 1077–1097. [Google Scholar] [CrossRef] - Dorninger, P.; Pfeifer, N. A comprehensive automated 3D approach for building extraction, reconstruction, and regularization from airborne laser scanning point clouds. Sensors
**2008**, 8, 7323–7343. [Google Scholar] [CrossRef] [PubMed] - He, Y.; Zhang, C.; Awrangjeb, M.; Fraser, C.S. Automated reconstruction of walls from airborne lidar data for complete 3D building modelling. In Proceedings of the International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Melbourne, Australia, 25 August–1 September 2012; pp. 115–120. [Google Scholar]
- Pu, S.; Vosselman, G. Knowledge based reconstruction of building models from terrestrial laser scanning data. ISPRS J. Photogramm. Remote Sens.
**2009**, 64, 575–584. [Google Scholar] [CrossRef] - Nguyen, A.; Le, B. 3D point cloud segmentation: A survey. In Proceedings of the 6th Conference on Robotics, Automation and Mechatronics (RAM), Manila, Philippines, 12–15 November 2013; pp. 225–230. [Google Scholar]
- Schnabel, R.; Wahl, R.; Klein, R. Efficient RANSAC for point-cloud shape detection. Comput. Graph. Forum
**2007**, 26, 214–226. [Google Scholar] [CrossRef] - Online Statistics Education: A Free Resource for Introductory Statistics. Available online: http://onlinestatbook.com/mobile/index.html (accessed on 11 May 2018).
- Rusu, R.B. PCL Tutorial. Available online: http://pointclouds.org/documentation/tutorials/statistical_outlier.php (accessed on 12 August 2016).
- Gross, H.; Thoennessen, U. Extraction of lines from laser point clouds. In Proceedings of the Symposium of ISPRS Commission III Photogrammetric Computer Vision, Bonn, Germany, 20–22 September 2006; Volume 36, pp. 86–91. [Google Scholar]
- Sampath, A.; Shan, J. Segmentation and Reconstruction of Polyhedral Building Roofs From Aerial Lidar Point Clouds. IEEE Trans. Geosci. Remote Sens.
**2010**, 48, 1554–1567. [Google Scholar] [CrossRef] - Rutzinger, M.; Rottensteiner, F.; Pfeifer, N. A Comparison of Evaluation Techniques for Building Extraction From Airborne Laser Scanning. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens.
**2009**, 2, 11–20. [Google Scholar] [CrossRef]

**Figure 3.**The first level of segmentation: (

**a**,

**c**) using cylinder primitive in dividing point clouds of the buildings, to (

**d**,

**f**) several parts; (

**b**) using cylinder primitive to (

**e**) segment point cloud of the footprint. (Units of axes are m).

**Figure 4.**Approximation of µ. (

**a**) The eigenvectors of an ellipse and a circle-shaped neighborhood around the centre point. (

**b**) The histogram of $1/\left({\lambda}_{3\text{}e}\text{}.\text{}{\lambda}_{2\text{}e}\right)$.

**Figure 5.**Approximating number of points of the planar segments for the efficient RANSAC. (

**a**) The histogram of cosine of the inclination angle of $\overrightarrow{{\lambda}_{1}}$. (

**b**–

**e**) Histograms of the inclination angle (in radian) of the footprint, roof, step roof and walls. (

**b’**,

**c’**,

**d’**,

**e’**) corresponding marked points in clusters (units of axes are m).

**Figure 6.**Planarity. (

**a**,

**c**) for the point clouds of the roof (5–8, 7–19), the walls (1–4, 1–7), and (

**b**,

**d**) for the point clouds of the grounds. (

**a**,

**b**) point clouds of building 1, and (

**c**,

**d**) point clouds of building 2.

**Figure 7.**Superimposing the constructed models on point clouds of (

**a**) building 1 and (

**b**) building 2. (Units of axes are m).

**Figure 8.**The reconstructed B-rep models. (

**a**,

**d**) superimposing the reconstructed buildings on the point clouds, (

**b**,

**e**) the reconstructed roofs, (

**c**,

**f**) the reconstructed footprints (green) besides the reference data (blue). (Units of axes are m).

Parameter | Value (s) |
---|---|

Maximum distance to the primitive (m) | 0.15 |

Maximum normal deviation (degree) | 10–15 |

Sampling resolution (m) | 0.1 |

Overlooking probability (%) | 85–90 |

Minimum points per primitive | Equation (5) |

Parameter | Values |
---|---|

Maximum distance to the primitive (m) | 0.01–0.05 |

Maximum normal deviation (degree) | 1–8 |

Sampling resolution (m) | 0.03–0.1 |

Overlooking probability (%) | 95–99 |

Minimum points per primitive | ${n}_{{p}_{i}}$ |

**Table 3.**Performance evaluation of the proposed segmentation method compared with the original RANSAC shape detection method.

$\mathbf{Completness}$ | $\mathbf{Correctness}$ | $\mathbf{Quality}$ | ||||
---|---|---|---|---|---|---|

Building1 | Building2 | Building1 | Building2 | Building1 | Building2 | |

the original efficient RANSAC method | 88% | 86% | 88% | 83% | 78% | 73% |

Our method | 90% | 88% | 100% | 97% | 90% | 86% |

Building 1 Accuracy (m) | Building 2 Accuracy (m) | ||||
---|---|---|---|---|---|

Horizontal | Vertical | Horizontal | Vertical | ||

Roof | Vertices of ridges | 0.11 | 0.11 | 0.09 | 0.10 |

Vertices of outer edges of eaves | 0.24 | 0.23 | 0.35 | 0.37 | |

Vertices of step roofs | - | - | 0.28 | 0.32 | |

Vertices calculated from intersection | 0.14 | 0.21 | 0.21 | 0.23 | |

Vertices of cylinders | - | - | 0.29 | 0.30 | |

Vertices of footprints | 0.19 | 0.17 | 0.28 | 0.24 |

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Malihi, S.; Valadan Zoej, M.J.; Hahn, M.
Large-Scale Accurate Reconstruction of Buildings Employing Point Clouds Generated from UAV Imagery. *Remote Sens.* **2018**, *10*, 1148.
https://doi.org/10.3390/rs10071148

**AMA Style**

Malihi S, Valadan Zoej MJ, Hahn M.
Large-Scale Accurate Reconstruction of Buildings Employing Point Clouds Generated from UAV Imagery. *Remote Sensing*. 2018; 10(7):1148.
https://doi.org/10.3390/rs10071148

**Chicago/Turabian Style**

Malihi, Shirin, Mohammad Javad Valadan Zoej, and Michael Hahn.
2018. "Large-Scale Accurate Reconstruction of Buildings Employing Point Clouds Generated from UAV Imagery" *Remote Sensing* 10, no. 7: 1148.
https://doi.org/10.3390/rs10071148