O-Snap: Optimization-Based Snapping for Modeling Architecture




Information

  • ACM Transactions on Graphics, 32(1):6:1-6:15
  • Publication Type: Journal Paper with Conference Talk
  • Year: 2013
  • Keywords: surface reconstruction, geometric optimization, interactive modeling

Abstract

In this article, we introduce a novel reconstruction and modeling pipeline to create polygonal models from unstructured point clouds. We propose an automatic polygonal reconstruction that can then be interactively refined by the user. An initial model is automatically created by extracting a set of RANSAC-based locally fitted planar primitives along with their boundary polygons, and then searching for local adjacency relations among parts of the polygons. The extracted set of adjacency relations is enforced to snap polygon elements together, while simultaneously fitting to the input point cloud and ensuring the planarity of the polygons. This optimization-based snapping algorithm may also be interleaved with user interaction. This allows the user to sketch modifications with coarse and loose 2D strokes, as the exact alignment of the polygons is automatically performed by the snapping. The generated models are coarse, offer simple editing possibilities by design, and are suitable for interactive 3D applications like games, virtual environments, etc. The main innovation in our approach lies in the tight coupling between interactive input and automatic optimization, as well as in an algorithm that robustly discovers the set of adjacency relations.

Result Images


Results from five point cloud data sets (generated from photos) shown from top to bottom: town hall, old church (Photos courtesy of Rainer Brechtken), mountain house (Photos courtesy of Sinha et al. [2008]), castle-P19 (Photos courtesy of Strecha et al. [2008]) and playhouse (Photos courtesy of Sinha et al. [2008]). Left to right: input point cloud, initial automatic reconstruction, refined model, final model with advanced details added, final textured model (with the photos simply back-projected onto the model).

Videos

Loading the player ...

O-Snap main video, demonstrating the features of our new system.

Loading the player ...

O-Snap applications video, demonstrating further use cases.

Acknowledgements

We wish to express our thanks to the reviewers for their insightful comments, and to the user study participants for their efforts and valuable feedback. We also thank Mariette Yvinec for providing us with the output mesh of their algorithm for the Church of Lans le Villard, and Evelyn Sutterlüti for the upfolding of the paper model of town hall. For providing data sets, we thank Sudipta Sinha (mountain house, playhouse), Rainer Brechtken (old church) and the Aim@Shape repository (Church of Lans le Villard). This work was partially supported by the Austrian Research Promotion Agency (FFG) through the FIT-IT project "Terapoints", project no. 825842. The competence center VRVis is funded by BMVIT, BMWFJ, and City of Vienna (ZIT) within the scope of COMET - Competence Centers for Excellent Technologies. The program COMET is managed by FFG.

Downloads

Paper:

Paper (Draft): [49 MB]
  This is the authors preprint. The definitive version of the paper will be available at
the ACM Digital Library - ACM Transactions on Graphics (TOG).


Additional Material:

Main Video (mp4): [123 MB]
Applications Video (mp4): [109 MB]
Result Images (png, zipped): [18 MB]

Bibtex

@article{arikan-2013-osn,
  title =      "O-Snap: Optimization-Based Snapping for Modeling Architecture",
  author =     "Murat Arikan and Michael Schw{\"a}rzler and Simon Fl{\"o}ry and Michael Wimmer and Stefan Maierhofer",
  year =       "2013",
  abstract =   "In this paper, we introduce a novel reconstruction and
               modeling pipeline to create polygonal models from
               unstructured point clouds. We propose an automatic polygonal
               reconstruction that can then be interactively refined by the
               user. An initial model is automatically created by
               extracting a set of RANSAC-based locally fitted planar
               primitives along with their boundary polygons, and then
               searching for local adjacency relations among parts of the
               polygons. The extracted set of adjacency relations is
               enforced to snap polygon elements together, while
               simultaneously fitting to the input point cloud and ensuring
               the planarity of the polygons. This optimization-based
               snapping algorithm may also be interleaved with user
               interaction. This allows the user to sketch modifications
               with coarse and loose 2D strokes, as the exact alignment of
               the polygons is automatically performed by the snapping. The
               generated models are coarse, offer simple editing
               possibilities by design and are suitable for interactive 3D
               applications like games, virtual environments etc. The main
               innovation in our approach lies in the tight coupling
               between interactive input and automatic optimization, as
               well as in an algorithm that robustly discovers the set of
               adjacency relations.",
  pages =      "6:1--6:15",
  month =      jan,
  numberr =    "1",
  event =      "ACM SIGGRAPH 2013 Annual Conference",
  journal =    "ACM Transactions on Graphics",
  volume =     "32",
  location =   "Anaheim, CA, USA",
  keywords =   "surface reconstruction, geometric optimization, interactive
               modeling",
  URL =        "http://www.cg.tuwien.ac.at/research/publications/2013/arikan-2013-osn/",
}
|  maintained by michael schwärzler   |   schwaerzler -at- vrvis.at   |   last update: 27.02.2013   |