Audrey Ledoux

Perceptual color hit-or-miss transform: application to dermatological image processing

The hit-or-miss transform is a mathematical morphological processing designed to find objects in image. Its extension to grayscale domain is not unique, but Barat’s method is the most appropriate to find specific objects with bandwidth in space and color evolutions. This tolerance is possible with the use of non-flat structuring elements. In this paper, a color extension of the Barat’s method is presented. For this purpose, a new mathematical morphology method, based on the concept of convergence in the CIELAB space and where the definition of non-flat structuring elements is possible, is used. A comparison with the existing approaches in the literature is done. Results are given and commented on synthetic and real images.

Toward a full-band texture features for spectral images

Facing the increasing number of multi and hyperspectral image acquisitions, in particular for medical and industrial applications, we need accurate features to analyse and assess the content complexity in a metrological way. In this paper, we explore an original way to compute texture features for spectral images in a full-band and vector process. To do it, we developed a dedicated approach for Mathematical Morphology using distance function. Thanks to this, we extend the classical mathematical morphology to spectral images. We show in this paper the scientific construction and preliminary results.

How to specify or identify the most accurate multispectral distance function for mathematical morphology

We developed a distance-based formalism adapted to colour mathematical morphology. To extend the morphological operators to multispectral domain, a suitable distance function between two spectra has to be specified and selected. In order to compare the behaviour of different distance functions, we apply linear transformations on basic spectrum. Then we compare the impact of the distance function choice during morphological process on real spectral images.

Color and multiscale texture features from vectorial mathematical morphology

In this paper, we explore an original way to compute texture features for color images in a vector process. Using a dedicated approach for color ordering, we produce a complete framework for color mathematical morphology adapted to human visual system characteristics. Then, morphological multiscale texture features are defined. To understand the texture feature behavior, we present the feature response to basic images variations. Finally, we compare the texture feature performance in front of a classical classification task using Outex database.

Vector texture features for colour images: Construction and elements of comparison

In this paper, we explore an original way to compute texture features for colour images in a vector process. To do it, we used a dedicated approach for colour mathematical morphology using distance function. We show in this paper the scientific construction of morphological spectra. We compare the obtained morphological spectra using different colour mathematical morphology methods existing in from the literature. First classification results are given using a fractal image database.

Renormalisation couleur d’une image de peau pour l’imagerie dermatologique

RÉSUMÉ. Les lésions dermatologiques sont suivies sur plusieurs mois ou années par le dermatologue, sans système d’acquisition d’images robuste et reproductible dans le temps. Compte tenu des contraintes d’usage du dermatologue, nous voulons construire un système d’acquisition faiblement contraint. Malheureusement, des variations de couleurs apparaissent et doivent être corrigées. Dans ce travail, nous comparons et analysons quelques modèles basiques de correction utilisés dans la littérature et nous montrons que la normalisation couleur est une transformation non linéaire dans des cadres d’acquisition insuffisamment contraints.

Problématique de l’idempotence pour les images couleurs et multivaluées

RÉSUMÉ. Le filtrage morphologique avancé nécessite que certaines propriétés soient validées avant d’être utilisées : idempotence et croissance. Si la preuve de ces propriétés est directe sur des objets unidimensionnels, il n’en est pas de même sur des objets multivalués. Ce travail porte sur la validation de la propriété d’idempotence des opérateurs d’ouverture et de fermeture couleur. Pour l’ordonnancement basé distance, nous montrons pourquoi cette propriété n’est pas vérifiée, nous proposons ensuite une modification résolvant ce problème puis un cadre générique d’écriture des fonctions d’ordonnancement vectorielles.