The image analysis techniques presented in the présent thesis have been developed as part of a European projeet dedicated to the development of an automatic membrane protein crystallization pipeline. A large number of samples is simultaneously produced and assessed by transmission electron microscope (TEM) screening. Automating this fast step implicates an on-fine analysis of acquired images to assure the microscope control by selecting the regions to be observed at high magnification and identify the components for specimen characterization.The observation of the sample at medium magnification provides the information that is essential to characterize the success of the 2D crystallization. The resulting objects, and especially the artificial membranes, are identifiable at this scale. These latter present only a few characteristic signatures, appearing in an extremely noisy context with gray-level fluctuations. Moreover they are practically transparent to electrons yielding low contrast. This thesis presents an ensemble of image processing techniques to analyze medium magnification images (5-15 nm/pixel). The original contribution of this work lies in: i) a statistical evaluation of contours by measuring the correlation between gray-levels of neighbouring pixels to the contour and a gradient signal for over-segmentation reduction, ii) the recognition of foreground entities of the image and iii) an initial study for their classification. This chain has been already tested on-line on a prototype and is currently evaluated.

Les techniques d'analyse des images présentées dans cette thèse sont élaborées dans le cadre du Projet Européen dédié au développement d'une plateforme automatique pour l'évaluation de la cristallisation des protéines membranaires. Un grand nombre d'échantillons est simultanément produit et évalué par microscopie électronique en transmission (MET). Pour rendre cette tache automatique, une analyse en ligne des images acquises est indispensable ; des régions d'intérêt essentielles pour le pilotage du microscope sont progressivement sélectionnées afin d'évaluer les cristaux de protéines a fort grossissement.L'observation de l'échantillon à moyen grossissement fournit des informations nécessaires pour la caractérisation du succès de la cristallisation 2D. Les objets résultants, et en particulier le gris. De plus, ces membranes sont pratiquement transparentes aux électrons et donc, apparaissent faiblement contrastées.Cette thèse présente un ensemble de techniques de traitement d'images pour leur analyse à moyen grossissement (5-15 nm/pixel). La contribution originale de ce travail est située sur i) une évaluation statistique des contours en mesurant la corrélation entre les niveaux de gris proche du contour et un filtre de gradient pour réduire le sur segmentation, ii) la reconnaissance des objets de l'image, iii) une étude préliminaire de classification. Cette chaîne est en cours de validation sur un prototype. / The image analysis techniques presented in the présent thesis have been developed as part of a European projeet dedicated to the development of an automatic membrane protein crystallization pipeline. A large number of samples is simultaneously produced and assessed by transmission electron microscope (TEM) screening. Automating this fast step implicates an on-fine analysis of acquired images to assure the microscope control by selecting the regions to be observed at high magnification and identify the components for specimen characterization.The observation of the sample at medium magnification provides the information that is essential to characterize the success of the 2D crystallization. The resulting objects, and especially the artificial membranes, are identifiable at this scale. These latter present only a few characteristic signatures, appearing in an extremely noisy context with gray-level fluctuations. Moreover they are practically transparent to electrons yielding low contrast. This thesis presents an ensemble of image processing techniques to analyze medium magnification images (5-15 nm/pixel). The original contribution of this work lies in: i) a statistical evaluation of contours by measuring the correlation between gray-levels of neighbouring pixels to the contour and a gradient signal for over-segmentation reduction, ii) the recognition of foreground entities of the image and iii) an initial study for their classification. This chain has been already tested on-line on a prototype and is currently evaluated.