Color monogenic wavelet representation based on a tensor-like use of the riesz transform: application to image coding - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2012

Color monogenic wavelet representation based on a tensor-like use of the riesz transform: application to image coding

Raphaël Soulard
SIC
Philippe Carré
SIC

Résumé

We propose a new extension of monogenic analysis to multi-valued signals like color images. This generalization is based on an analogy between the Riesz transform and structure tensors and takes advantage of the well defined vector differential geometry. Our color wavelet transform is non-marginal and its coefficients - separated into amplitude, phase, orientation and local color axis - have interesting physical interpretation in terms of local energy, contour model, and colorimetric features. An image coding application is proposed as a practical study.
Fichier principal
Vignette du fichier
Soulard2012b_agacse16color_preprint.pdf (720.32 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-00740130 , version 1 (09-10-2012)

Identifiants

  • HAL Id : hal-00740130 , version 1

Citer

Raphaël Soulard, Philippe Carré. Color monogenic wavelet representation based on a tensor-like use of the riesz transform: application to image coding. Applied Geometric Algebras in Computer Science and Engineering (AGACSE 2012), Jul 2012, La Rochelle, France. Paper 16. ⟨hal-00740130⟩
175 Consultations
101 Téléchargements

Partager

Gmail Facebook X LinkedIn More