Bayesian 3D Modeling from Images using Multiple Depth Maps

Pau Gargallo 1 Peter Sturm 1
1 MOVI - Modeling, localization, recognition and interpretation in computer vision
GRAVIR - IMAG - Graphisme, Vision et Robotique, Inria Grenoble - Rhône-Alpes, CNRS - Centre National de la Recherche Scientifique : FR71
Abstract : This paper addresses the problem of reconstructing the geometry and color of a Lambertian scene, given some fully calibrated images acquired with wide baselines. In order to completely model the input data, we propose to represent the scene as a set of colored depth maps, one per input image. We formulate the problem as a Bayesian MAP problem which leads to an energy minimization method. Hidden visibility variables are used to deal with occlusion, reflections and outliers. The main contributions of this work are: a prior for the visibility variables that treats the geometric occlusions; and a prior for the multiple depth maps model that smoothes and merges the depth maps while enabling discontinuities. Real world examples showing the efficiency and limitations of the approach are presented.
Document type :
Conference papers
Complete list of metadatas

Cited literature [19 references]  Display  Hide  Download

https://hal.inria.fr/inria-00524394
Contributor : Peter Sturm <>
Submitted on : Wednesday, May 25, 2011 - 9:58:53 AM
Last modification on : Wednesday, April 11, 2018 - 1:52:13 AM
Long-term archiving on: Friday, August 26, 2011 - 2:20:29 AM

File

GargalloSturm-cvpr05.pdf
Files produced by the author(s)

Identifiers

Collections

CNRS | INRIA | IMAG | UGA

Citation

Pau Gargallo, Peter Sturm. Bayesian 3D Modeling from Images using Multiple Depth Maps. IEEE Workshop on Motion and Video Computing, Jun 2005, San Diego, United States. pp.885-891, ⟨10.1109/CVPR.2005.84⟩. ⟨inria-00524394⟩

Share

Metrics

Record views

423

Files downloads

488