Vous êtes ici : Accueil » Kiosque » Annonce

Identification

Identifiant: 
Mot de passe : 

Mot de passe oublié ?
Détails d'identification oubliés ?

Annonce

4 avril 2018

Phd Proposal on "Texture Features for Hyperspectral Image Analysis"


Catégorie : Doctorant


Texture Features for Hyperspectral Image Analysis

The XLIM laboratory(Poitiers-France) and the Colour and Norwegian Visual Computing Laboratory (Gjøvik-Noway) are looking for candidates for a Ph.D position on "Texture Features for Hyperspectral Image Analysis".

The objectives of this PhD is to propose the first metrological framework for texture assessment of non-uniformity in hyperspectral images, generally called the “textured aspect”. To assess the accuracy and other metrological constraints, the validation will be processed using real samples with known optical properties (that will be provided by the CRCC-Research Center for the Conservation of Collections – French National Museum of National History who are partners of this project). To separate the uncertainties due to the real samples and the uncertainties due to the digital processing, artificial images will also be synthesized using mathematical models. In particular, a methodology will be developed for linking the texture assessment to the complexity assessment in the spectral case, following the work developed previously for the colour domain (Richard, et al., 2016).

 

Texture Features for Hyperspectral Image Analysis

Direction de thèse : Noël Richard, Christine Fernandez-Maloigne

Co-direction de thèse : Jon Y. Hardeberg (NTNU, Gjøvik), Christine Andraud (MNHN)

Mots clefs : Hyperspectral, Texture, Feature, Metrology, Complexity, Non-Linear Processing

Contact :

Noël Richard, noel.richard@univ-poitiers.fr

 

Dans cette rubrique

(c) GdR 720 ISIS - CNRS - 2011-2018.