Outils d'évaluation de la qualité d'un paramétrage de propriétés visuelles : cas des textures couleur

Abstract : Nowadays, sensory properties of materials are subject to growing attention both in hedonic point of view than utilitarian. Hence, the formulation of the foundations of an instrumental metrological approach that will allow the characterization of visual similarities between textures belonging to the same type becomes a challenge of the research activities in the domain of perception. The specific objective of this work is to link an instrumental approach of metrology of the assessment of visual textures with metrology approach based on softcopy experiment performed by human observers. The experiment consisted in ranking of isochromatic colored textured images according to the visual contrast. The collected data have been analyzed by means of two statistical methods. First, data have been analyzed by fitting a fixed effects factorial model to the midranks statistics. The second method relies on an extension of Mallows-Bradley-Terry (MBT) model for ranking data with ties. Maximum likelihood estimation of MBT models parameters was performed by means of MM algorithms and MCMC procedures. A likelihood ratio test procedure was proposed in order to choose between the two following hypotheses : perceptive discernibility of the textures and non perceptive discernibility of the textures. The computation of the test statistics was carried out using a Monte Carlo procedure. As a result of the two statistical analyses, the visual contrast appears to be a sensory continuum that can be quantified by building a numerical sensory scale.
Complete list of metadatas

Cited literature [134 references]  Display  Hide  Download

Contributor : Sylvie Beuhorry-Sassus <>
Submitted on : Friday, February 26, 2010 - 3:32:36 PM
Last modification on : Friday, March 29, 2019 - 10:46:11 AM
Long-term archiving on : Friday, June 18, 2010 - 8:26:55 PM


  • HAL Id : tel-00460261, version 1



Amadou Sawadogo. Outils d'évaluation de la qualité d'un paramétrage de propriétés visuelles : cas des textures couleur. Mathématiques [math]. Université de Pau et des Pays de l'Adour, 2009. Français. ⟨tel-00460261⟩



Record views


Files downloads