Éditeur(s) : HAL CCSD
Résumé : International audience
In this paper, we use a multifractal approach based on the computation of two spectrums for image analysis and texture segmentation problems. The two spectrums are the Legendre Spectrum, determined by classical methods, and the Large Deviation Spectrum, determined by kernel density estimation. We propose a way for the fusion of these two spectrums to improve textured image segmentation results. An unsupervised k-means is used as clustering approach for the texture classification. The algorithm is applied on mosaic image built using IKONOS images and various natural textures from the Brodatz album. The segmentation obtained with our approach gives better results than the application of each spectrum separately.
CIARP Proceedings
Cancun, Mexico

hal-00602268
https://hal.archives-ouvertes.fr/hal-00602268

Éditeur(s) : HAL CCSD
Résumé : En traitement d'image, avant toute exploitation de résultats d'une segmentation d'image, il est souvent nécessaire de réaliser une opération de fusion. Le but est de créer une partition formée de composantes connexes, modélisant les objets présents dans l'image. La segmentation initiale des images est souvent obtenue par l'exploitation d'attributs de bas-niveau (couleur, texture, forme...). La théorie Gestalt permet de prendre en compte des informations d'un niveau supérieur qui s'appuient sur l'organisation spatiale des objets et groupement d'objets initialement segment'es. Dans cet article, nous proposons une méthode de fusion de régions. L'originalité des travaux repose sur l'ajout d'un nouveau critère qui, associé aux critères Gestalt, permet de prendre en compte les objets jugés d'intérêt par le photographe, c'est-'a-dire les objets aux contours nets. Les régions associées aux objets d'intérêts seront alors préservées alors que les régions floues ou aux contours flous, situées dans l'arrière-plan, auront tendance 'a fusionner entre elles.
Journée du GdR Mathématique de l'Optimisation et Application (MOA)
Saint-François, Guadeloupe

hal-00706538
https://hal.archives-ouvertes.fr/hal-00706538

Éditeur(s) : HAL CCSD
Résumé : International audience
This paper deals with image processing. This study takes place in a segmentation process based on texture analysis. We use the multifractal ap-proach to characterize the textures. More precisely we study a particular multi-fractal spectrum called the large deviation spectrum. We consider two statistical methods to numerically compute this spectrum. The resulting spectrum, com-puted by both methods over an image, is a one dimension spectrum. In the scope of this article, we extend these methods in order to obtain a two dimen-sions spectrum which could be assimilated to an image. This 2D spectrum al-lows a local characterization of the image singularities while a 1D spectrum is a global characterization. Moreover, the computation of the spectrum requires the use of a measure. We introduce here a pre processing based on the gradient to improve the measure. We show results on both synthetic and real world images. Finally, we remark that the resulting 2D spectrum is close to the resulting im-age of an edge detection process while edge detection using one dimension spectrum requires post processing methods. This statement will be used for fu-ture works.
SITIS Proceedings
Hammamet, Tunisia

hal-00602273
https://hal.archives-ouvertes.fr/hal-00602273

Éditeur(s) : HAL CCSD
Résumé : International audience
This paper deals with a comparison of different colour space in order to improve high resolution images classification. The background of this study is the measure of the agriculture impact on the environment in islander context. Biodiversity is particularly sensitive and relevant in such areas and the follow-up of the forest front is a way to ensure its preservation. Very high resolution satellite images are used such as QuickBird and IKONOS scenes. In order to segment the images into forest and agriculture areas, we characterize both ground covers with colour and texture features. A classical unsupervised classifier is then used to obtain labelled areas. As features are computed on coloured images, we can wonder if the colour space choice is relevant. This study has been made considering more than fourteen colour spaces (RGB, YUV, Lab, YIQ, YCrCs, XYZ, CMY, LMS, HSL, KLT, IHS, I1I2I3, HSV, HSI, etc.) and shows the visual and quantitative superiority of IHS on all others. For conciseness reasons, results only show RGB, I1I2I3 and IHS colour spaces.
SPIE Remote Sensing Proceedings
Cardiff, United Kingdom

hal-00602277
https://hal.archives-ouvertes.fr/hal-00602277

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
We carried out detailed petrographic, major and trace element geochemical, microstructural and FTIR analyses on eight characteristic ultramafic xenoliths from Nyos and Barombi Mbo Lakes in the continental sector of the Cameroon Volcanic Line (CVL). The studied xenoliths are spinel lherzolites showing lithologies similar to the other xenoliths reported previously along the CVL. They have protogranular and porphyroclastic textures. One of the Barombi xenolith contains amphibole, which had not been previously reported in this locality. Amphibole is common in the Nyos xenoliths suite. Peridotite xenoliths from both localities show some chemical heterogeneity, but Barombi xenoliths generally are less depleted in basaltic elements with respect to Nyos xenoliths. Trace element compositions of Nyos spinel lherzolites show a moderately depleted initial (premetasomatic) composition and variable enrichment in REE. Evidence for both modal and cryptic metasomatism is present in Nyos xenoliths. Rare earth element patterns of clinopyroxene suggest that interaction between mafic melts and the upper mantle occurred beneath the Nyos locality. Barombi Mbo xenoliths, on the other hand, record a small degree of partial melting. The Barombi Mbo xenoliths have weak, dominantly orthorhombic olivine crystal preferred orientations, whereas Nyos ones have strong axial-[010] patterns, which may have formed in response to transpression. Nominally anhydrous mantle minerals (NAMs) of the Barombi Mbo xenoliths show generally higher bulk concentrations of ‘water’ (70–127 ppm) than Nyos xenoliths (32–81 ppm). The Barombi Mbo xenoliths could originate from a juvenile segment of the lithospheric mantle, which had been originally part of the asthenosphere. It became a part of the lithosphere in response to thermal relaxation following the extension, forming a weakly deformed lower lithospheric mantle region along the CVL. The Nyos xenoliths, however, represent a shallow lithospheric mantle bearing imprints of several depletion and enrichment events probably prior or following the extension (at ∼30 Ma).
ISSN: 0899-5362

hal-01236039
https://hal.archives-ouvertes.fr/hal-01236039
DOI : 10.1016/j.jafrearsci.2015.07.006

Éditeur(s) : HAL CCSD
Résumé : International audience
Half away between edge detection and object recognition this study aim to characterize singularities in order to guide this two phases. The caracterization is obtained with a generalization of the Lipschitz exponents .for complex structures. This new characterization is called " maxima chain " and integrates the notion of value and spatial repartition of the maxima. The maxima chains are obtained with a multiscale decomposition using a Diadic Discrete Wavelet Transform (DDWT). We use the maxima of the gradient image computed with the wavelet details. The local maxima detect the position of irregular structures (edges) in the image (Canny algorithm). Then we characterize each maxima using their evolution across the different scales. A classification is made using the maxima chains. We integrate the classification to eliminate unwanted maxima (from noise or other objects) to drive the search of closed contours. In fact, there is a totally different signature from one maxima to another depending on the parameters of the objects : nature, shape, size, gray scale, texture, ...
RFIA Proceedings
Toulouse, France

hal-00602265
https://hal.archives-ouvertes.fr/hal-00602265