Adaptive Color Transfer With Relaxed Optimal Transport

Julien Rabin 1 Sira Ferradans 2 Nicolas Papadakis 3
1 Equipe Image - Laboratoire GREYC - UMR6072
GREYC - Groupe de Recherche en Informatique, Image, Automatique et Instrumentation de Caen
Abstract : This paper tackles the problem of color transfer between images using discrete optimal transportation. Methods that only match discrete image color palettes through bijective assignments [1], without considering the spatial pixel location such as [2,3], suffer from several limitations such as spatial inconsistencies or noise amplification, so that additional post-processing must be considered in practice [1,4]. To tackle these issues, we here propose a variational model that directly deals with the regularity of the transport map and the spatial consistency of the reconstruction. Our approach is based on the relaxed and regularized discrete optimal transport method of [5]. We extend this model in order to take into account the spatial distribution of colors within the image domain and only relax locally the bijective assignment constraint, when required by the data. We rely on a spatio-color clustering framework to ensure fast computation while preserving the quality of the color transfer. Finally, we present some experiments on real images that demonstrate the capacity of our model to adapt itself to the considered data. [1] Piti e, F., Kokaram, A.C., Dahyot, R.: Automated colour grading using colour distribution transfer. Computer Vision and Image Understanding 107 (2007) 123-137 [2] Papadakis, N., Provenzi, E., Caselles, V.: A variational model for histogram transfer of color images. IEEE Transactions on Image Processing 20 (2011) 1682-1695 [3] Rabin, J., Peyr e, G.: Wasserstein regularization of imaging problem. In: IEEE International Conderence on Image Processing (ICIP'11). (2011) 1541-1544 [4] Rabin, J., Delon, J. and Gousseau, Y.: Regularization of transportation maps for color and contrast transfer, In: IEEE International Conderence on Image Processing (ICIP'10). (2010) 1933-1936 [5] Ferradans, A., Papadakis, N., Rabin, J., Peyré, G. and Aujol, J-F. Regularized discrere optimal transport. In: International Conference on Scale Space and Variational Methods in Computer Vision (2013) 1-12
Type de document :
Communication dans un congrès
IEEE international Conference on Image Processing, Oct 2014, Paris, France. pp.5, 2014
Liste complète des métadonnées

Littérature citée [21 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01002830
Contributeur : Julien Rabin <>
Soumis le : vendredi 6 juin 2014 - 18:04:44
Dernière modification le : lundi 25 septembre 2017 - 16:08:14
Document(s) archivé(s) le : samedi 6 septembre 2014 - 12:36:17

Fichier

icip14_HAL.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01002830, version 1

Collections

Citation

Julien Rabin, Sira Ferradans, Nicolas Papadakis. Adaptive Color Transfer With Relaxed Optimal Transport. IEEE international Conference on Image Processing, Oct 2014, Paris, France. pp.5, 2014. 〈hal-01002830〉

Partager

Métriques

Consultations de
la notice

584

Téléchargements du document

527