Template Assembly for Detailed Urban Reconstruction

Liangliang Nan, Caigui Jiang, Bernard Ghanem, Peter Wonka

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

We propose a new framework to reconstruct building details by automatically assembling 3D templates on coarse textured building models. In a preprocessing step, we generate an initial coarse model to approximate a point cloud computed using Structure from Motion and Multi View Stereo, and we model a set of 3D templates of facade details. Next, we optimize the initial coarse model to enforce consistency between geometry and appearance (texture images). Then, building details are reconstructed by assembling templates on the textured faces of the coarse model. The 3D templates are automatically chosen and located by our optimization-based template assembly algorithm that balances image matching and structural regularity. In the results, we demonstrate how our framework can enrich the details of coarse models using various data sets.

Original languageEnglish (US)
Pages (from-to)217-228
Number of pages12
JournalComputer Graphics Forum
Volume34
Issue number2
DOIs
StatePublished - May 1 2015
Externally publishedYes

Fingerprint

Image matching
Facades
Textures
Geometry

ASJC Scopus subject areas

  • Computer Networks and Communications

Cite this

Nan, L., Jiang, C., Ghanem, B., & Wonka, P. (2015). Template Assembly for Detailed Urban Reconstruction. Computer Graphics Forum, 34(2), 217-228. https://doi.org/10.1111/cgf.12554

Template Assembly for Detailed Urban Reconstruction. / Nan, Liangliang; Jiang, Caigui; Ghanem, Bernard; Wonka, Peter.

In: Computer Graphics Forum, Vol. 34, No. 2, 01.05.2015, p. 217-228.

Research output: Contribution to journalArticle

Nan, L, Jiang, C, Ghanem, B & Wonka, P 2015, 'Template Assembly for Detailed Urban Reconstruction', Computer Graphics Forum, vol. 34, no. 2, pp. 217-228. https://doi.org/10.1111/cgf.12554
Nan, Liangliang ; Jiang, Caigui ; Ghanem, Bernard ; Wonka, Peter. / Template Assembly for Detailed Urban Reconstruction. In: Computer Graphics Forum. 2015 ; Vol. 34, No. 2. pp. 217-228.
@article{e1ff92b0083046ccb38d76dc26b69f1f,
title = "Template Assembly for Detailed Urban Reconstruction",
abstract = "We propose a new framework to reconstruct building details by automatically assembling 3D templates on coarse textured building models. In a preprocessing step, we generate an initial coarse model to approximate a point cloud computed using Structure from Motion and Multi View Stereo, and we model a set of 3D templates of facade details. Next, we optimize the initial coarse model to enforce consistency between geometry and appearance (texture images). Then, building details are reconstructed by assembling templates on the textured faces of the coarse model. The 3D templates are automatically chosen and located by our optimization-based template assembly algorithm that balances image matching and structural regularity. In the results, we demonstrate how our framework can enrich the details of coarse models using various data sets.",
author = "Liangliang Nan and Caigui Jiang and Bernard Ghanem and Peter Wonka",
year = "2015",
month = "5",
day = "1",
doi = "10.1111/cgf.12554",
language = "English (US)",
volume = "34",
pages = "217--228",
journal = "Computer Graphics Forum",
issn = "0167-7055",
publisher = "Wiley-Blackwell",
number = "2",

}

TY - JOUR

T1 - Template Assembly for Detailed Urban Reconstruction

AU - Nan, Liangliang

AU - Jiang, Caigui

AU - Ghanem, Bernard

AU - Wonka, Peter

PY - 2015/5/1

Y1 - 2015/5/1

N2 - We propose a new framework to reconstruct building details by automatically assembling 3D templates on coarse textured building models. In a preprocessing step, we generate an initial coarse model to approximate a point cloud computed using Structure from Motion and Multi View Stereo, and we model a set of 3D templates of facade details. Next, we optimize the initial coarse model to enforce consistency between geometry and appearance (texture images). Then, building details are reconstructed by assembling templates on the textured faces of the coarse model. The 3D templates are automatically chosen and located by our optimization-based template assembly algorithm that balances image matching and structural regularity. In the results, we demonstrate how our framework can enrich the details of coarse models using various data sets.

AB - We propose a new framework to reconstruct building details by automatically assembling 3D templates on coarse textured building models. In a preprocessing step, we generate an initial coarse model to approximate a point cloud computed using Structure from Motion and Multi View Stereo, and we model a set of 3D templates of facade details. Next, we optimize the initial coarse model to enforce consistency between geometry and appearance (texture images). Then, building details are reconstructed by assembling templates on the textured faces of the coarse model. The 3D templates are automatically chosen and located by our optimization-based template assembly algorithm that balances image matching and structural regularity. In the results, we demonstrate how our framework can enrich the details of coarse models using various data sets.

UR - http://www.scopus.com/inward/record.url?scp=84932118845&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84932118845&partnerID=8YFLogxK

U2 - 10.1111/cgf.12554

DO - 10.1111/cgf.12554

M3 - Article

VL - 34

SP - 217

EP - 228

JO - Computer Graphics Forum

JF - Computer Graphics Forum

SN - 0167-7055

IS - 2

ER -