Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform

Lars Bodum, Jens Overby, Erik Kjems, Peer Møller Ilsøe

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearch

373 Downloads (Pure)

Abstract

Urban environments of modern cities are described digitally in large public databases and datasets of e.g. laser scanning and ortho photos. These data sets are not necessarily linked to each other, except trough their geometry attributes (coordinates), which are mutually displaced and have a low degree of details. However, it is possible to create virtual 3D models of buildings, by processing these data. Roof polygons are generated using airborne laser scanning of 1x1 meter grid and ground plans (footprints) extracted from technical feature maps. An effective algorithm is used for fixing the mutual displacement of these datasets. The well known Hough Transform is extended to 3D for extracting planes from the point cloud. Generally speaking, planes are rejected if the clusters of points on these planes, do not span a considerable area. Furthermore, it is assumed that valid planes are close to parallel to one of the ground plan lines. Points corresponding to each valid plane are subtracted from the original point cloud, and the Hough Transform is performed on the remainder. By this approach, the disturbing influence of already evaluated points, is avoided. Small variations of the roof surface might lead to multiple slightly differing planes. Such planes are detected and merged. Intersecting planes are identified, and a polygon mesh of the roof is constructed. Due to the low precision of the laser scanning, a rule-based postprocessing of the roof is applied before adding the walls.
Original languageEnglish
Title of host publicationISPRS proceedings
Number of pages6
Volumexxxv-B3
PublisherThe Organising Committee of the XXth International Congress for Photogrammetry and Remote Sensing
Publication date2004
Pages1-6
Publication statusPublished - 2004
EventXX ISPRS Congress, Commission III - Istanbul, Turkey
Duration: 12 Jul 200423 Jul 2004

Conference

ConferenceXX ISPRS Congress, Commission III
CountryTurkey
CityIstanbul
Period12/07/200423/07/2004

Fingerprint

transform
roof
laser
polygon
footprint
trough
geometry

Keywords

  • Point Cloud
  • Laser scanning
  • GIS
  • Algorithms
  • Building
  • Geometry
  • Reconstruction
  • Virtual Reality

Cite this

Bodum, L., Overby, J., Kjems, E., & Ilsøe, P. M. (2004). Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform. In ISPRS proceedings (Vol. xxxv-B3, pp. 1-6). The Organising Committee of the XXth International Congress for Photogrammetry and Remote Sensing.
Bodum, Lars ; Overby, Jens ; Kjems, Erik ; Ilsøe, Peer Møller. / Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform. ISPRS proceedings. Vol. xxxv-B3 The Organising Committee of the XXth International Congress for Photogrammetry and Remote Sensing, 2004. pp. 1-6
@inproceedings{aee46230004411dab4d5000ea68e967b,
title = "Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform",
abstract = "Urban environments of modern cities are described digitally in large public databases and datasets of e.g. laser scanning and ortho photos. These data sets are not necessarily linked to each other, except trough their geometry attributes (coordinates), which are mutually displaced and have a low degree of details. However, it is possible to create virtual 3D models of buildings, by processing these data. Roof polygons are generated using airborne laser scanning of 1x1 meter grid and ground plans (footprints) extracted from technical feature maps. An effective algorithm is used for fixing the mutual displacement of these datasets. The well known Hough Transform is extended to 3D for extracting planes from the point cloud. Generally speaking, planes are rejected if the clusters of points on these planes, do not span a considerable area. Furthermore, it is assumed that valid planes are close to parallel to one of the ground plan lines. Points corresponding to each valid plane are subtracted from the original point cloud, and the Hough Transform is performed on the remainder. By this approach, the disturbing influence of already evaluated points, is avoided. Small variations of the roof surface might lead to multiple slightly differing planes. Such planes are detected and merged. Intersecting planes are identified, and a polygon mesh of the roof is constructed. Due to the low precision of the laser scanning, a rule-based postprocessing of the roof is applied before adding the walls.",
keywords = "Point Cloud, Laser scanning, GIS, Algorithms, Building, Geometry, Reconstruction, Virtual Reality",
author = "Lars Bodum and Jens Overby and Erik Kjems and Ils{\o}e, {Peer M{\o}ller}",
note = "Findes ogs{\aa} p{\aa} dvd. ISSN ; 1682-1777",
year = "2004",
language = "English",
volume = "xxxv-B3",
pages = "1--6",
booktitle = "ISPRS proceedings",
publisher = "The Organising Committee of the XXth International Congress for Photogrammetry and Remote Sensing",

}

Bodum, L, Overby, J, Kjems, E & Ilsøe, PM 2004, Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform. in ISPRS proceedings. vol. xxxv-B3, The Organising Committee of the XXth International Congress for Photogrammetry and Remote Sensing, pp. 1-6, Istanbul, Turkey, 12/07/2004.

Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform. / Bodum, Lars; Overby, Jens; Kjems, Erik; Ilsøe, Peer Møller.

ISPRS proceedings. Vol. xxxv-B3 The Organising Committee of the XXth International Congress for Photogrammetry and Remote Sensing, 2004. p. 1-6.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearch

TY - GEN

T1 - Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform

AU - Bodum, Lars

AU - Overby, Jens

AU - Kjems, Erik

AU - Ilsøe, Peer Møller

N1 - Findes også på dvd. ISSN ; 1682-1777

PY - 2004

Y1 - 2004

N2 - Urban environments of modern cities are described digitally in large public databases and datasets of e.g. laser scanning and ortho photos. These data sets are not necessarily linked to each other, except trough their geometry attributes (coordinates), which are mutually displaced and have a low degree of details. However, it is possible to create virtual 3D models of buildings, by processing these data. Roof polygons are generated using airborne laser scanning of 1x1 meter grid and ground plans (footprints) extracted from technical feature maps. An effective algorithm is used for fixing the mutual displacement of these datasets. The well known Hough Transform is extended to 3D for extracting planes from the point cloud. Generally speaking, planes are rejected if the clusters of points on these planes, do not span a considerable area. Furthermore, it is assumed that valid planes are close to parallel to one of the ground plan lines. Points corresponding to each valid plane are subtracted from the original point cloud, and the Hough Transform is performed on the remainder. By this approach, the disturbing influence of already evaluated points, is avoided. Small variations of the roof surface might lead to multiple slightly differing planes. Such planes are detected and merged. Intersecting planes are identified, and a polygon mesh of the roof is constructed. Due to the low precision of the laser scanning, a rule-based postprocessing of the roof is applied before adding the walls.

AB - Urban environments of modern cities are described digitally in large public databases and datasets of e.g. laser scanning and ortho photos. These data sets are not necessarily linked to each other, except trough their geometry attributes (coordinates), which are mutually displaced and have a low degree of details. However, it is possible to create virtual 3D models of buildings, by processing these data. Roof polygons are generated using airborne laser scanning of 1x1 meter grid and ground plans (footprints) extracted from technical feature maps. An effective algorithm is used for fixing the mutual displacement of these datasets. The well known Hough Transform is extended to 3D for extracting planes from the point cloud. Generally speaking, planes are rejected if the clusters of points on these planes, do not span a considerable area. Furthermore, it is assumed that valid planes are close to parallel to one of the ground plan lines. Points corresponding to each valid plane are subtracted from the original point cloud, and the Hough Transform is performed on the remainder. By this approach, the disturbing influence of already evaluated points, is avoided. Small variations of the roof surface might lead to multiple slightly differing planes. Such planes are detected and merged. Intersecting planes are identified, and a polygon mesh of the roof is constructed. Due to the low precision of the laser scanning, a rule-based postprocessing of the roof is applied before adding the walls.

KW - Point Cloud

KW - Laser scanning

KW - GIS

KW - Algorithms

KW - Building

KW - Geometry

KW - Reconstruction

KW - Virtual Reality

M3 - Article in proceeding

VL - xxxv-B3

SP - 1

EP - 6

BT - ISPRS proceedings

PB - The Organising Committee of the XXth International Congress for Photogrammetry and Remote Sensing

ER -

Bodum L, Overby J, Kjems E, Ilsøe PM. Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform. In ISPRS proceedings. Vol. xxxv-B3. The Organising Committee of the XXth International Congress for Photogrammetry and Remote Sensing. 2004. p. 1-6