Fast global stereo matching via energy pyramid minimization

B Conejo 1, 2, 3 S Leprince 3 F Ayoub 3 Jean-Philippe Avouac 3
1 IMAGINE [Marne-la-Vallée]
2 LIGM - Laboratoire d'Informatique Gaspard-Monge
3 Division of Geological and Planetary Sciences [Pasadena]
Abstract : We define a global matching framework based on energy pyramid, the Global Matching via Energy Pyramid (GM-EP) algorithm, which estimates the disparity map from a single stereo-pair by solving an energy minimization problem. We efficiently address this minimization by globally optimizing a coarse to fine sequence of sparse Conditional Random Fields (CRF) directly defined on the energy. This global discrete optimization approach guarantees that at each scale we obtain a near optimal solution, and we demonstrate its superiority over state of the art image pyramid approaches through application to real stereo-pairs. We conclude that multiscale approaches should be build on energy pyramids rather than on image pyramids.
Keywords : Photogrammetry Global Matching Multiresolution Algorithms Estimation DEM/DTM
Type de document :
Communication dans un congrès
Photogrammetric Computer Vision - PCV2014, Sep 2014, Zurich, Switzerland. 3, pp.41 - 48, 2014, 〈10.5194/isprsannals-II-3-41-2014〉
Liste complète des métadonnées

Littérature citée [33 références]  Voir  Masquer  Télécharger
https://hal-enpc.archives-ouvertes.fr/hal-01081593
Contributeur : Pascal Monasse <>
Soumis le : lundi 10 novembre 2014 - 09:22:24
Dernière modification le : mercredi 11 avril 2018 - 12:12:03
Document(s) archivé(s) le : mercredi 11 février 2015 - 15:23:14

Fichier

PCV14.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

ENPC | IMAGINE | LIGM_A3SI | LIGM | PARISTECH | UPEC-UPEM

Citation

B Conejo, S Leprince, F Ayoub, Jean-Philippe Avouac. Fast global stereo matching via energy pyramid minimization. Photogrammetric Computer Vision - PCV2014, Sep 2014, Zurich, Switzerland. 3, pp.41 - 48, 2014, 〈10.5194/isprsannals-II-3-41-2014〉. 〈hal-01081593〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

295

Téléchargements de fichiers

409

Contact

support.ccsd.cnrs.fr
hal.support@ccsd.cnrs.fr
Logo CCSD