Mémoire (M.Sc.T.)--Unviversité du Québec à Chicoutimi, 2001. / Document électronique également accessible en format PDF. CaQCU

Existem milhares de espécies de serpentes no mundo, muitas com padrões distintos e intricados. Esta diversidade se torna um problema para usuários que precisam criar texturas de pele de serpente para aplicar em modelos 3D, pois a dificuldade em criar estes padrões complexos é considerável. Nós primeiramente propomos uma categorização de padrões de pele de serpentes levando em conta suas características visuais. Então apresentamos um modelo procedural capaz de sintetizar uma vasta gama de textura de padrões de pele de serpentes. O modelo usa processamento de imagem simples (tal como sintetizar bolinhas e listras) bem como autômatos celulares e geradores de ruído para criar texturas realistas para usar em renderizadores modernos. Nossos resultados mostram boa similaridade visual com pele de serpentes reais. As texturas resultantes podem ser usadas não apenas em computação gráfica, mas também em educação sobre serpentes e suas características visuais. Nós também realizamos testes com usuários para avaliar a usabilidade de nossa ferramenta. O escore da Escala de Usabilidade do Sistema foi de 85:8, sugerindo uma ferramenta de texturização altamente efetiva. / There are thousands of snake species in the world, many with intricate and distinct skin patterns. This diversity becomes a problem for users who need to create snake skin textures to apply on 3D models, as the difficulty for creating such complex patterns is considerable. We first propose a categorization of snake skin patterns considering their visual characteristics. We then present a procedural model capable of synthesizing a wide range of texture skin patterns from snakes. The model uses simple image processing (such as synthesizing spots and stripes) as well as cellular automata and noise generators to create realistic textures for use in a modern renderer. Our results show good visual similarity with real skin found in snakes. The resulting textures can be used not only for computer graphics texturing, but also in education about snakes and their visual characteristics. We have also performed a user study to assess the usability of our tool. The score from the System Usability Scale was 85:8, suggesting a highly effective texturing tool.

Dans cette thèse, des matériaux microporeux carbonés et décorés par des nanoparticules métalliques ont été synthétisés afin d’étudier l’effet de la décoration par les nanoparticules métalliques sur leurs propriétés d’adsorption d’hydrogène. L’utilisation de saccharose, de tannins de mimosa ou encore de paille de riz était en parfaite cohérence avec une démarche de respect de l’environnement et de développement durable. Le saccharose ainsi que les tannins permettent, via un prétraitement hydrothermal, d’obtenir des hydrochars qui sont ensuite activés par voie chimique ou physique. Ces synthèses ont donné accès, en fonction de paramètres de synthèse, à des matériaux à la porosité contrôlée dont les propriétés d’adsorption d’hydrogène à température ambiante ont ensuite été évaluées. En outre, l’intérêt majeur de cette thèse était de proposer un couplage entre les propriétés d’adsorption d’hydrogène, des basses pressions jusqu’à des pressions relativement élevées, et les propriétés texturales modélisées à l’aide d’outils théoriques, tels que la DFT, appliquée au fluide inhomogènes (Density Functionnal Theory). L’utilisation de dioxyde de carbone supercritique, de synthèses hydrothermales et des différents précurseurs a permis d’obtenir des matériaux dopés au nickel, au silicium ou à l’azote sans utilisation de solvant organiques préjudiciables pour la santé. Finalement, la décorrélation des différentes contributions de d’adsorption a permis de détecter et de quantifier les fractions d’hydrogène adsorbées et engagées dans divers processus tels que la physisorption, la chimisorption, le spillover et de détecter les phénomènes physisorption polarisée / In this PhD work, microporous carbonaceous materials decorated with metallic nanoparticles were developed in order to study the effect of nanodispersed metals on hydrogen sorption properties. Bio-sourced precursors such as sucrose, mimosa tannins or even rice straw have been used in order to promote eco-friendly, sustainable and innovating ways of synthesis. The use of hydrothermal pretreatment on tannins or sucrose allowed obtaining hydrochars that have been chemically or physically activated. Depending on synthesis parameters, these syntheses allowed to produce materials with controlled porosity. Then, the hydrogen adsorption properties of these materials were evaluated at ambient temperature as well as their textural properties. The main interest of this work was to propose a coupling between hydrogen sorption properties, from low to high pressure, and the textural properties. The textural properties were modelized using theoretical tools such as the DFT (Density Functionnal Theory). The use the different precursors and of decoration steps using supercritical carbon dioxide or hydrothermal synthesis, during which ones metallic or nitrogen containing precursors were used, allowed to obtain nickel, nitrogen or silicon doped materials without using harmful organic solvents. Finally, the decorrelation of the different adsorption contributions gave the possibility to detect and quantify the hydrogen fractions engaged in different processes such as physisorption, chemisorption and spillover. Polarized or enhanced physisorption was also detected using the same methods

Existem milhares de espécies de serpentes no mundo, muitas com padrões distintos e intricados. Esta diversidade se torna um problema para usuários que precisam criar texturas de pele de serpente para aplicar em modelos 3D, pois a dificuldade em criar estes padrões complexos é considerável. Nós primeiramente propomos uma categorização de padrões de pele de serpentes levando em conta suas características visuais. Então apresentamos um modelo procedural capaz de sintetizar uma vasta gama de textura de padrões de pele de serpentes. O modelo usa processamento de imagem simples (tal como sintetizar bolinhas e listras) bem como autômatos celulares e geradores de ruído para criar texturas realistas para usar em renderizadores modernos. Nossos resultados mostram boa similaridade visual com pele de serpentes reais. As texturas resultantes podem ser usadas não apenas em computação gráfica, mas também em educação sobre serpentes e suas características visuais. Nós também realizamos testes com usuários para avaliar a usabilidade de nossa ferramenta. O escore da Escala de Usabilidade do Sistema foi de 85:8, sugerindo uma ferramenta de texturização altamente efetiva. / There are thousands of snake species in the world, many with intricate and distinct skin patterns. This diversity becomes a problem for users who need to create snake skin textures to apply on 3D models, as the difficulty for creating such complex patterns is considerable. We first propose a categorization of snake skin patterns considering their visual characteristics. We then present a procedural model capable of synthesizing a wide range of texture skin patterns from snakes. The model uses simple image processing (such as synthesizing spots and stripes) as well as cellular automata and noise generators to create realistic textures for use in a modern renderer. Our results show good visual similarity with real skin found in snakes. The resulting textures can be used not only for computer graphics texturing, but also in education about snakes and their visual characteristics. We have also performed a user study to assess the usability of our tool. The score from the System Usability Scale was 85:8, suggesting a highly effective texturing tool.

abstract: Texture analysis plays an important role in applications like automated pattern inspection, image and video compression, content-based image retrieval, remote-sensing, medical imaging and document processing, to name a few. Texture Structure Analysis is the process of studying the structure present in the textures. This structure can be expressed in terms of perceived regularity. Our human visual system (HVS) uses the perceived regularity as one of the important pre-attentive cues in low-level image understanding. Similar to the HVS, image processing and computer vision systems can make fast and efficient decisions if they can quantify this regularity automatically. In this work, the problem of quantifying the degree of perceived regularity when looking at an arbitrary texture is introduced and addressed. One key contribution of this work is in proposing an objective no-reference perceptual texture regularity metric based on visual saliency. Other key contributions include an adaptive texture synthesis method based on texture regularity, and a low-complexity reduced-reference visual quality metric for assessing the quality of synthesized textures. In order to use the best performing visual attention model on textures, the performance of the most popular visual attention models to predict the visual saliency on textures is evaluated. Since there is no publicly available database with ground-truth saliency maps on images with exclusive texture content, a new eye-tracking database is systematically built. Using the Visual Saliency Map (VSM) generated by the best visual attention model, the proposed texture regularity metric is computed. The proposed metric is based on the observation that VSM characteristics differ between textures of differing regularity. The proposed texture regularity metric is based on two texture regularity scores, namely a textural similarity score and a spatial distribution score. In order to evaluate the performance of the proposed regularity metric, a texture regularity database called RegTEX, is built as a part of this work. It is shown through subjective testing that the proposed metric has a strong correlation with the Mean Opinion Score (MOS) for the perceived regularity of textures. The proposed method is also shown to be robust to geometric and photometric transformations and outperforms some of the popular texture regularity metrics in predicting the perceived regularity. The impact of the proposed metric to improve the performance of many image-processing applications is also presented. The influence of the perceived texture regularity on the perceptual quality of synthesized textures is demonstrated through building a synthesized textures database named SynTEX. It is shown through subjective testing that textures with different degrees of perceived regularities exhibit different degrees of vulnerability to artifacts resulting from different texture synthesis approaches. This work also proposes an algorithm for adaptively selecting the appropriate texture synthesis method based on the perceived regularity of the original texture. A reduced-reference texture quality metric for texture synthesis is also proposed as part of this work. The metric is based on the change in perceived regularity and the change in perceived granularity between the original and the synthesized textures. The perceived granularity is quantified through a new granularity metric that is proposed in this work. It is shown through subjective testing that the proposed quality metric, using just 2 parameters, has a strong correlation with the MOS for the fidelity of synthesized textures and outperforms the state-of-the-art full-reference quality metrics on 3 different texture databases. Finally, the ability of the proposed regularity metric in predicting the perceived degradation of textures due to compression and blur artifacts is also established. / Dissertation/Thesis / Ph.D. Electrical Engineering 2014

Au cours de ces dernières années, les matrices de covariance ont montré leur intérêt dans de nombreuses applications en traitement du signal et de l'image.Les travaux présentés dans cette thèse se concentrent sur l'utilisation de ces matrices comme descripteurs pour la classification. Dans ce contexte, des algorithmes robustes de classification sont proposés en développant les aspects suivants.Tout d'abord, des estimateurs robustes de la matrice de covariance sont utilisés afin de réduire l'impact des observations aberrantes. Puis, les distributions Riemannienne Gaussienne et de Laplace, ainsi que leur extension au cas des modèles de mélange, sont considérés pour la modélisation des matrices de covariance.Les algorithmes de type k-moyennes et d'espérance-maximisation sont étendus au cas Riemannien pour l'estimation de paramètres de ces lois : poids, centroïdes et paramètres de dispersion. De plus, un nouvel estimateur du centroïde est proposé en s'appuyant sur la théorie des M-estimateurs : l'estimateur de Huber. En outre,des descripteurs appelés vecteurs Riemannien de Fisher sont introduits afin de modéliser les images non-stationnaires. Enfin, un test d'hypothèse basé sur la distance géodésique est introduit pour réguler la probabilité de fausse alarme du classifieur.Toutes ces contributions sont validées en classification d'images de texture, de signaux du cerveau, et d'images polarimétriques radar simulées et réelles. / In the recent years, covariance matrices have demonstrated their interestin a wide variety of applications in signal and image processing. The workpresented in this thesis focuses on the use of covariance matrices as signatures forrobust classification. In this context, a robust classification workflow is proposed,resulting in the following contributions.First, robust covariance matrix estimators are used to reduce the impact of outlierobservations, during the estimation process. Second, the Riemannian Gaussianand Laplace distributions as well as their mixture model are considered to representthe observed covariance matrices. The k-means and expectation maximization algorithmsare then extended to the Riemannian case to estimate their parameters, thatare the mixture's weight, the central covariance matrix and the dispersion. Next,a new centroid estimator, called the Huber's centroid, is introduced based on thetheory of M-estimators. Further on, a new local descriptor named the RiemannianFisher vector is introduced to model non-stationary images. Moreover, a statisticalhypothesis test is introduced based on the geodesic distance to regulate the classification false alarm rate. In the end, the proposed methods are evaluated in thecontext of texture image classification, brain decoding, simulated and real PolSARimage classification.

L’Hypertrophie Ventriculaire Gauche (HVG) est actuellement mise en évidence par échographie. Cet examen fournit des informations anatomo-fonctionnelles mais ne permet pas de déterminer l’étiologie des HVG, ce qui engendre de graves erreurs de diagnostic et de prise en charge thérapeutique. Les HVG sont classiquement séparées en 2 catégories :1. pathologies hypertrophiques induites par la modification structurelle et fonctionnelle des cardiomyocites qui tend à compenser des insuffisances cardiaques liées par exemple à des problèmes d’hypertension artérielle, de rétrécissement aortique ou de CardioMyopathies Hypertrophiques sarcomériques ;2. pathologies infiltratives correspondant au dépôt de protéines dans la matriceextracellulaire principalement dues à différentes formes d’amyloses cardiaquesNotre hypothèse est que les différents mécanismes physiopathologiques (hypertrophique ou infiltratif) pourraient se traduire dans l’image par des propriétés spécifiques du speckle échographique. Nous avons donc développé un travail d’analyse de la texture de ces images afin de discriminer les différentes HVG.Dans cette étude, la base de données de 4795 images est divisée en une base d’apprentissage de 3770 images et une base de test de 1025 images. L’analyse de texture des images est faite par les ondelettes de Gabor avec 8 orientations, 7 tailles et 5 niveaux de décomposition. Ensuite, les caractéristiques statistiques de premier et deuxième ordre sont extraites des images. Le nombre des caractéristiques est réduit pour la base d’apprentissage en appliquant l’Analyse en Composantes Principales (ACP) suivie par l’analyse discriminante linéaire (ADL) pour une séparation supervisée des classes. Les caractéristiques extraites pour la base de test sont projetées sur les vecteurs propres sélectionnés au cours de l’apprentissage. L’ADL est appliquée à ce niveau pour la classification des données du test et la qualité de cette classification est évaluée. Les résultats obtenus sont bons (qualité totale de classification de 95,51%) et sont suivis d’une étape de cross-validation afin de vérifier la robustesse de notre méthode. A cette étape, les bases de données de l’apprentissage et du test sont mélangées et 50 combinaisons différentes sont évaluées. La même méthode décrite précédemment est appliquée. La cross-validation montre une variation de la qualité de classification (entre 30% et 99.96%) probablement due à l’hétérogénéité des caractéristiques texturelles pour les patients d’une même classe que l’on peut expliquer par des degrés différents d’avancement dans la pathologie.Ces travaux montrent qu’une analyse de texture des images échocardiographiques peut permettre de déterminer des bio-marqueurs aptes à discriminer différentes cardiopathies qui s’expriment par une HVG. Ce résultat peut avoir des retombées très importantes dans la détection précoce des amyloses cardiaques, maladies engendrant un fort taux de mortalité souvent dû à un retard de diagnostic et prise en charge des patients par un centre expert / Left-Ventricular Hypertrophy (LVH) is currently detected through echocardiography. The latter imaging modality provides anatomical and functional information. However, it does not allow the determination of the HVG etiology. This can, in turn, lead to dangerous errors in the diagnosis and treatment planning of the disease. LVH pathologies are separated into two categories:- Hypertrophic pathology caused by the structural and functional modification of cardiomyocytes that lead to cardiac failure related, for example, to arterial hypertension problems, aortic narrowing or sarcomeric hypertrophic cardiomyopathies.- Infiltrative pathologies corresponding to protein deposits on the extracellular matrix, mainly due to different forms of cardiac amyloidosisOur hypothesis is that different physiopathological mechanisms (hypertrophic or infiltrative) can be translated in the image through properties specific to echographic speckle. We have therefore developed the work of texture analysis of such images in order to discriminate the different types of LVH.In this study, the database of 4795 images is divided into a learning database of 3770 images and another testing database of 1025 images. The textural analysis of these images is done using Gabor wavelets with 8 orientations, 7 sizes and 5 decomposition levels. Next, the statistical characteristics of first and second orders are extracted from the filtered images. The number of characteristics is reduced for the learning database by applying Principal Component Analysis (PCA) followed by Linear Discriminant Analysis (LDA) for a supervised separation of the classes. The extracted characteristics for the test database are projected on the eigenvectors selected in the learning step. LDA is applied at this level for the test data classification, and the quality of this classification is evaluated. The obtained results are good (total classification quality of 95.51%). A step of cross-validation follows in order to verify the robustness of our method. At this stage, the learning and testing databases are mixed, and 50 different combinations are evaluated. The same method described previously is then applied. The cross-validation shows a variation in the classification quality (between 30% and 99.96%) probably due to the heterogeneity of the texture characteristics for the patients of the same class explained by different disease advancement stages.This work shows that the textural analysis of echocardiographic images can permit the determination of bio-markers suitable to discriminate different LVH cardiopathies. Our results can have a very important impact on the early detection of cardiac amyloidosis, a pathology causing a considerable rate of mortality often due to a belated diagnosis and support by the centers of expertise

Image inpainting is the process of recovering missing or deteriorated data within the digital images and videos in a plausible way. It has become an important topic in the area of image processing, which leads to the understanding of the textural and structural information within the images. Image inpainting has many different applications, such as image/video restoration, text/object removal, texture synthesis, and transmission error concealment. In recent years, many algorithms have been developed to solve the image inpainting problem, which can be roughly grouped into four categories, partial differential equation-based inpainting, exemplar-based inpainting, transform domain inpainting, and hybrid image inpainting. However, the existing algorithms do not work well when the missing region to be inpainted is large, and when there are textural and structural information needed to be recovered. To address this inpainting problem, we propose multiple algorithms, 1) perceptually aware image inpainting based on the perceptual-fidelity aware mean squared error metric, 2) image inpainting using nonlocal texture matching and nonlinear filtering, and 3) multiresolution exemplar-based image inpainting. The experimental results show that our proposed algorithms outperform other existing algorithms with respect to both qualitative analysis and observer studies when inpainting the missing regions of images.

La plupart des méthodes projectives (Didier ANZIEU, Catherine CHABERT, 1961) ont comme support de travail la vue. Le processus haptique est-il suffisamment discriminatif pour valider la fiabilité de ce test sous ce mode ? Un étalonnage différencié respecte-t-il la valeur donnée à ce test par H. Rorschach ? Le sujet, en perdant la vue, va devoir mobiliser ou trouver des nouvelles ressources physiques et psychiques pour devenir autonome et poursuivre sa vie. Tout en lui est effracté : sa construction narcissique, sa représentation du monde et son altérité dans le regard de l’autre. Le Rorschach permet d’aborder la richesse psychique d’un sujet. Il nous donne à voir les potentialités psychiques du sujet dans sa relation à lui-même et aux autres. Or, il fait appel à une activité visuelle qui organise un matériel informel. C’est une activité imageante qui fait passer du perceptif au projectif, du perçu au vécu…Le toucher, dans certaines situations, peut être un relais de perception, de projection et donner à voir les représentations du sujet. Cette recherche peut permettre d’envisager une autre manière d’étalonner ce test afin de pouvoir le quantifier, le qualifier en changeant par exemple l’item « couleur » par l’item texture…Dans cette étude, il nous faut vérifier que cela n’enlève pas la valeur que donnait H. Rorschach à ce test, à cet item mais au contraire l’ouvre à une autre dimension clinique, celle d’une clinique spécifique du sujet mal ou non-voyant. En respectant un protocole spécifique, une telle passation permettrait une médiation pour le sujet en situation de handicap visuel vers une « meilleure » ré-élaboration de son projet de vie. Par ailleurs, cette étude pourrait ouvrir d’autres approches sur le dessin en relief et permettre notamment à toute une partie de la population vieillissante d’utiliser cette modalité dans la vie quotidienne, pour de nouveaux repères. / This research follows a study already begun in Master 1; it deals with the haptic process adapted to 3-dimensional Rorschach-plates. Is the haptic process sufficiently discriminating to validate the reliability of this test under this mode ? Does differentiated calibration respect the values given to this test by H. Rorschach ? After loss of sight, a subject is going to have to mobilize or find new physical and psychic resources to become autonomous and go on living. Everything in that person is broken: their narcissistic construction, their representation of the world and their differences in the eye of others. The Rorschach test allows us to approach a subject’s psychic wealth. It shows us the psychic potential of a subject in relation to himself and to others. Yet the Rorschach test calls on visual activity to organize an informal material. It is an activity based on images which leads one from the perceptive to the projective, from perceived to experienced. In certain situations, touch can be an intermediary for perception, for projection and can show how the subject is represented. Through this research, we can consider calibrating this test in another way in order to measure quantity and quality by changing “colour” for ‘texture’, for example. In this study, it is essential to verify that this does not remove the value given by H. Rorschach to this item in this test, but on the contrary opens it to other clinical dimensions, that specific of the blind or partially-sighted.Moreover, this study could open up new approaches to 3-dimensional drawings and allow a great part of the ageing population to use this method in daily life, to find new bearings. For instance, using certain textures to find one’s way around the flat by 3-dimensional pictograms.

La texture de l'albumen de blé tendre (Triticum aestivum) est une propriété importante du grain qui affecte son comportement au fractionnement, ainsi que la qualité des produits finis. Elle est définie par la dureté et la vitrosité qui sont deux propriétés différentes du blé. La dureté est reliée à l'adhésion entre les granules d'amidon et la matrice protéique (principaux constituants de l'albumen) qui est contrôlée génétiquement à travers l'état des puroindolines (sauvages ou mutées). La vitrosité est associée à la porosité de l'albumen et varie principalement en fonction des conditions de culture. Le principal objectif de ce travail de recherche est de contribuer à la compréhension du lien entre la texture de l'albumen et le comportement des grains à la mouture, en fonction de la génétique (puroindolines) et des conditions de culture (porosité). Des grains issus de lignées quasi-isogéniques, ne différant que par l'état allélique de la puroindoline b (Pinb-D1a pour les soft et Pinb-D1b pour les hard) et cultivés dans deux endroits différents (grains vitreux et farineux) ont été spécifiquement sélectionnés pour distinguer le rôle respectif des puroindolines et de la porosité sur la microstructure de l'albumen et sur les propriétés mécaniques à l'échelle de l'albumen et d'une population de grains. Cette analyse multi-échelles a permis de préciser que l'effet de la porosité de l'albumen domine sur la microstructure pour les grains farineux, indépendemment de la génétique, et que l'adhésion amidon-protéines est également impliquée dans le cas des grains vitreux. La porosité et l'adhésion entre les constituants de l'albumen sont tous deux responsables des différences observées de propriétés mécaniques, de comportement des grains à la mouture et de propriétés de la farine. A l'échelle nanométrique, nous avons réussi à mesurer les propriétés mécaniques locales de l'amidon et du gluten en associant une méthode originale utilisant la microscopie à force atomique (AFM) avec un modèle de tribologie. Pour la première fois, nous avons montré que la dureté de l'amidon est quatre fois plus élevée que celle du gluten, indépendamment de leur origine génétique. Nous avons aussi montré pour la première fois que les propriétés mécaniques de l'interface entre amidon et protéines étaient dépendantes de l'origine génétique des blés, et plus précisément des puroindolines. Ces résultats confirment que la dureté du grain n'est pas liée aux propriétés mécaniques des constituants du blé mais plutôt à l'interface amidon-protéine. Ces données ont été intégrées par la suite dans un modèle numérique qui permet de prédire le comportement mécanique global des échantillons en fonction du degré d'adhésion entre amidon et protéines, l'assemblage granulaire des particules d'amidon et la teneur en protéines. Ce modèle a permis de mettre en évidence l'effet de l'assemblage des granules d'amidon sur les propriétés mécaniques, qui a été jusqu'ici négligé dans l'évaluation de la vitrosité. / Endosperm texture is one of the most important grain properties for fractionation behaviour and end-use quality of common wheat (Triticum aestivum). Hardness and vitreousness are distinct grain properties that are both responsible of grain texture. Hardness is related to the starch and protein (endosperm main components) adhesion that is genetically controlled depending on the wild or mutated puroindolines. Vitreousness is associated to endosperm porosity and is mainly affected by the growing conditions. The principal aim of this PhD work is to contribute understanding the link between endosperm texture and the grain milling behaviour, depending on the wheat genetic background (puroindolines) and growth conditions (porosity). Near-isogenic lines differing only by the puroindoline b allelic state (Pinb-D1a for soft and Pinb-D1b for hard) and grown in different locations (vitreous and mealy kernels) were specifically selected to analyse the endosperm microstructure and mechanical properties at the kernel and grain population scales. The multi-scale analysis pointed out, in mealy grains, the dominant effect of porosity on the endosperm microstructure whatever the genetic background, and showed that the starch-protein adhesion is also involved in vitreous kernels. Both endosperm porosity and starch-protein adhesion are responsible for the distinct endosperm mechanical properties, milling behaviour and flour properties. At the nanoscale, an original nanoscratching method using Atomic Force Microscopy (AFM) was associated with a tribological model to measure the mechanical properties of wheat endosperm main components. For the first time, the hardness of starch was found four fold higher than that of gluten, whatever the genetic origin. Most importantly, the AFM methodology clearly revealed differences in the mechanical properties of starch-protein interface between hard and soft grains. These results confirm that grain hardness is related to the mechanical properties of the starch-protein interface and that the puroindolines nature is involved in these properties. The nano-mechanical properties of starch and proteins determined by AFM were integrated in the numerical modelling to predict wheat fractionation according to the starch-protein adhesion, the starch particles assembly and the protein content. The numerical model highlighted the effect of the starch granular assembly on wheat grains mechanical properties, which has not been taken into account before to evaluate the effect of vitreousness.