Tracking changes in buildings over time - Fully automated reconstruction and difference detection of 3D scan and BIM files

Martin Tamke, Mateusz Zwierzycki, Henrik Leander Evers, Sebastian Ochmann, Richard Vock und Raoul Wessel
In proceedings of the 34th eCAADe Conference, Oulu, Finland, pages 643-651, Aug. 2016
 

Abstract

Architectural and Engineering Communities are interested in the detection of differences between different representations of the same building. These can be the differences between the design and the as-built-state of a building, or the detection of changes that occur over time and that are documented by consecutive 3D scans. Current approaches for the detection of differences between 3D scans and 3D building models are however laborious and work only on the level of a building element. We demonstrate a novel highly automated workflow to detect differences between representations of the same building. We discuss the underlying tools and methods and the ways to communicate deviations and differences in an appropriate manner and evaluate our approach with a rich set of real world datasets.

Download: http://papers.cumincad.org/data/works/att/ecaade2016_147.pdf

Bilder

Bibtex

@INPROCEEDINGS{tamke-2016-tracking,
        author = {Tamke, Martin and Zwierzycki, Mateusz and Evers, Henrik Leander and Ochmann, Sebastian and Vock,
                  Richard and Wessel, Raoul},
         pages = {643--651},
         title = {Tracking changes in buildings over time - Fully automated reconstruction and difference detection of
                  3D scan and BIM files},
     booktitle = {the 34th eCAADe Conference},
        volume = {2},
          year = {2016},
         month = aug,
      location = {Oulu, Finland},
  howpublished = {To appear in Proceedings of the 34th eCAADe Conference},
      abstract = {Architectural and Engineering Communities are interested in the detection of
                  differences between different representations of the same building. These can be
                  the differences between the design and the as-built-state of a building, or the
                  detection of changes that occur over time and that are documented by consecutive
                  3D scans. Current approaches for the detection of differences between 3D scans
                  and 3D building models are however laborious and work only on the level of a
                  building element. We demonstrate a novel highly automated workflow to detect
                  differences between representations of the same building. We discuss the
                  underlying tools and methods and the ways to communicate deviations and
                  differences in an appropriate manner and evaluate our approach with a rich set
                  of real world datasets.}
}