ACCURATE DENSE STEREO MATCHING FOR ROAD SCENES

Oussama Zeglazi 1, * Mohammed Rziza 1 Aouatif Amine 1 Cédric Demonceaux 2
* Auteur correspondant
1 LRIT
LRIT - Laboratoire de Recherche Informatique et Télécommunications
Abstract : Stereo matching task is the core of applications linked to the intelligent vehicles. In this paper, we present a new variant function of the Census Transform (CT) which is more robust against radiometric changes in real road scenes. We demonstrate that the proposed cost function outperforms the conventional cost functions using the KITTI benchmark. The cost aggregation method is also updated for taking into account the edge information. This enables to improve significantly the aggregated costs especially within homogenous regions. The Winner-Takes-All (WTA) strategy is used to compute disparity values. To further eliminate the remainder matching ambiguities , a post-processing step is performed. Experiments were conducted on the new Middlebury 2 dataset, as well as on the real road traffic scenes of the KITTI database. Obtained disparity results have demonstrated that the proposed method is promising.
Type de document :
Communication dans un congrès
IEEE International Conference on Image Processing, Sep 2017, Beijing, China. IEEE International Conference on Image Processing, 2017, 〈http://2017.ieeeicip.org/index.asp〉
Liste complète des métadonnées

Littérature citée [20 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01592090
Contributeur : Cédric Demonceaux <>
Soumis le : vendredi 22 septembre 2017 - 15:40:36
Dernière modification le : mercredi 13 décembre 2017 - 14:20:33

Fichier

ICIP2017.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01592090, version 1

Collections

Citation

Oussama Zeglazi, Mohammed Rziza, Aouatif Amine, Cédric Demonceaux. ACCURATE DENSE STEREO MATCHING FOR ROAD SCENES. IEEE International Conference on Image Processing, Sep 2017, Beijing, China. IEEE International Conference on Image Processing, 2017, 〈http://2017.ieeeicip.org/index.asp〉. 〈hal-01592090〉

Partager

Métriques

Consultations de la notice

61

Téléchargements de fichiers

24