Factored Facade Acquisition using Symmetric Line Arrangements

Handle URI:
http://hdl.handle.net/10754/598308
Title:
Factored Facade Acquisition using Symmetric Line Arrangements
Authors:
Ceylan, Duygu; Mitra, Niloy J.; Li, Hao; Weise, Thibaut; Pauly, Mark
Abstract:
We introduce a novel framework for image-based 3D reconstruction of urban buildings based on symmetry priors. Starting from image-level edges, we generate a sparse and approximate set of consistent 3D lines. These lines are then used to simultaneously detect symmetric line arrangements while refining the estimated 3D model. Operating both on 2D image data and intermediate 3D feature representations, we perform iterative feature consolidation and effective outlier pruning, thus eliminating reconstruction artifacts arising from ambiguous or wrong stereo matches. We exploit non-local coherence of symmetric elements to generate precise model reconstructions, even in the presence of a significant amount of outlier image-edges arising from reflections, shadows, outlier objects, etc. We evaluate our algorithm on several challenging test scenarios, both synthetic and real. Beyond reconstruction, the extracted symmetry patterns are useful towards interactive and intuitive model manipulations.
Citation:
Ceylan D, Mitra NJ, Li H, Weise T, Pauly M (2012) Factored Facade Acquisition using Symmetric Line Arrangements. Computer Graphics Forum 31: 671–680. Available: http://dx.doi.org/10.1111/j.1467-8659.2012.03046.x.
Publisher:
Wiley-Blackwell
Journal:
Computer Graphics Forum
Issue Date:
May-2012
DOI:
10.1111/j.1467-8659.2012.03046.x
Type:
Article
ISSN:
0167-7055
Sponsors:
This research has been supported by the ERC Starting Grant 257453 COSYM, a KAUST visiting student grant, and the Marie Curie Career Integration Grant 303541.
Appears in Collections:
Publications Acknowledging KAUST Support

Full metadata record

DC FieldValue Language
dc.contributor.authorCeylan, Duyguen
dc.contributor.authorMitra, Niloy J.en
dc.contributor.authorLi, Haoen
dc.contributor.authorWeise, Thibauten
dc.contributor.authorPauly, Marken
dc.date.accessioned2016-02-25T13:18:26Zen
dc.date.available2016-02-25T13:18:26Zen
dc.date.issued2012-05en
dc.identifier.citationCeylan D, Mitra NJ, Li H, Weise T, Pauly M (2012) Factored Facade Acquisition using Symmetric Line Arrangements. Computer Graphics Forum 31: 671–680. Available: http://dx.doi.org/10.1111/j.1467-8659.2012.03046.x.en
dc.identifier.issn0167-7055en
dc.identifier.doi10.1111/j.1467-8659.2012.03046.xen
dc.identifier.urihttp://hdl.handle.net/10754/598308en
dc.description.abstractWe introduce a novel framework for image-based 3D reconstruction of urban buildings based on symmetry priors. Starting from image-level edges, we generate a sparse and approximate set of consistent 3D lines. These lines are then used to simultaneously detect symmetric line arrangements while refining the estimated 3D model. Operating both on 2D image data and intermediate 3D feature representations, we perform iterative feature consolidation and effective outlier pruning, thus eliminating reconstruction artifacts arising from ambiguous or wrong stereo matches. We exploit non-local coherence of symmetric elements to generate precise model reconstructions, even in the presence of a significant amount of outlier image-edges arising from reflections, shadows, outlier objects, etc. We evaluate our algorithm on several challenging test scenarios, both synthetic and real. Beyond reconstruction, the extracted symmetry patterns are useful towards interactive and intuitive model manipulations.en
dc.description.sponsorshipThis research has been supported by the ERC Starting Grant 257453 COSYM, a KAUST visiting student grant, and the Marie Curie Career Integration Grant 303541.en
dc.publisherWiley-Blackwellen
dc.titleFactored Facade Acquisition using Symmetric Line Arrangementsen
dc.typeArticleen
dc.identifier.journalComputer Graphics Forumen
dc.contributor.institutionEPFLen
dc.contributor.institutionUCLen
dc.contributor.institutionColumbia Universityen
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.