Dissertations / Theses on the topic 'Traitement d'images – Techniques numériques – Segmentation bayésienne'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Traitement d'images – Techniques numériques – Segmentation bayésienne.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Benboudjema, Dalila. "Champs de Markov triplets et segmentation bayésienne non supervisée d'images." Evry, Institut national des télécommunications, 2005. http://www.theses.fr/2005TELE0009.
Full textImage segmentation is a fundamental and yet difficult task in machine vision. Several models and approaches have been proposed, and the ones which have probably received considerable attention are hidden Markov fields (HMF) models. In such model the hidden field X which is assumed Markovian, must be estimated from the observed –or noisy- field Y. Such processing is possible because the distribution X conditional on the observed process Y remains markovian. This model has been generalized to the Pairwise Markov field (PMF) which offer similar processing and superior modelling capabilities. In this model we assume directly the markovianity of the couple (X,Y ). Afterwards, triplet Markov fields (TMF) which are the generalization of the PMF, have been proposed. In such model the distribution of the couple (X ,Y ) is the marginal distribution of a Markov field T = (X ,U,Y ) , where U is latent process. The aim of this thesis is to study the TMF models. Two original models are presented: the Evidential Markov field (EMF) allowing to model the evidential aspects of the prior information and the adapted triplet Markov field (ATMF), allowing to model the simultaneous presence of different stationarities in the class image. For the unsupervised processing, two original approaches of estimation the model’s parameters have been proposed. The first one is based on the stochastic gradient and the second one is based on the iterative conditional estimation (ICE) and the least square method, as well. The latter, have then been generalized to the non stationary images with non Gaussian correlated noise, which uses the Pearson system to find the natures of margins of the noise, which can vary with the class. Experiments indicate that the new models and related processing algorithms can improve the results obtained with the classical ones
Quelle, Hans-Christoph. "Segmentation bayesienne non supervisee en imagerie radar." Rennes 1, 1993. http://www.theses.fr/1993REN10012.
Full textMignotte, Max. "Segmentation d'images sonar par approche markovienne hiérarchique non supervisée et classification d'ombres portées par modèles statistiques." Brest, 1998. http://www.theses.fr/1998BRES2017.
Full textBricq, Stéphanie. "Segmentation d’images IRM anatomiques par inférence bayésienne multimodale et détection de lésions." Université Louis Pasteur (Strasbourg) (1971-2008), 2008. https://publication-theses.unistra.fr/public/theses_doctorat/2008/BRICQ_Stephanie_2008.pdf.
Full textMedical imaging provides a growing number of data. Automatic segmentation has become a fundamental step for quantitative analysis of these images in many brain diseases such as multiple sclerosis (MS). We focused our study on brain MRI segmentation and MS lesion detection. At first we proposed a method of brain tissue segmentation based on hidden Markov chains taking into account neighbourhood information. This method can also include prior information provided by a probabilistic atlas and takes into account the artefacts appearing on MR images. Then we extended this method to detect MS lesions thanks to a robust estimator and prior information provided by a probabilistic atlas. We have also developed a 3D MRI segmentation method based on statistical active contours to refine the lesion segmentation. The results were compared with other existing methods of segmentation, and with manual expert segmentations
Chapoulie, Alexandre. "Contributions aux méthodes de détection visuelle de fermeture de boucle et de segmentation topologique de l'environnement." Nice, 2012. http://www.theses.fr/2012NICE4055.
Full textIn the context of global localization and, more widely, in Simultaneous Localization and Mapping, it is mandatory to be able to detect if a robust comes to a previously visited place. It is the loop closure detection problem. Algorithms, in visual place recognition, usually allow detection in real-time, are robust to perceptual aliasing or even to dynamic objects. Those algorithms are often sensitive to the robot orientation involving an impossibility to detect a loop closure from a different point of view. In order to alleviate this drawback, panoramic or omnidirectional cameras are often used. We propose a more general representation of the environment with an ego-centric spherical view. Using these representation properties, we elaborate a loop closure detection algorithm that satisfies, in addition to other properties, robot orientation independence. The environment model is often a set of images taken at various moments, each image corresponding to a place. Existing methods cluster those images ion meaning places of the environment, the topological places, using the concept of covisibility of information between places. Our approach relies on the utilization of the environment structure. We hence define a topological place as having a structure which does not change, variation leading to a place change. The structure variations are detected with an efficient change-point detection algorithm
Rousson, Mikaël. "Cue integration and front evolution in image segmentation." Nice, 2004. http://www.theses.fr/2004NICE4100.
Full textAutomatic detection and selection of regions of interest is a key step in image understanding. In the literature, most segmentation approaches are restricted to a particular class of images. This limitation is due to the large variety of cues available to characterize a region of interest. Targeting particular applications, algorithms are centered on the from most relevant cue. The limiting factor to obtain a general algorithm is the large variety of cues available to characterize a region of interest. It can be gray-level, color, texture, shape, etc. . . In this thesis, we propose a general formulation able to deal with each one of these characteristics. Image intensity, color, texture, motion and prior shape knowledge are considered. For this purpose, a probabilistic inference is obtained from a Bayesian formulation of the segmentation problem. Then, reformulated as an energy minimization, the most probable image partition is obtained using front evolution techniques. Level-set functions are naturally introduced to represent the evolving fronts while region statistics are optimized in parallel. This framework can naturally handle scalar and vector-valued smooth images but more complex cues are also integrated. Texture and motion features, as well as prior shape knowledge are successively introduced. Complex medical images are considered in the last part with the case of diffusion magnetic resonance images which gives 3D probability density fields
Scherrer, Benoit. "Segmentation des tissus et structures sur les IRM cérébrales : agents markoviens locaux et coopératifs et formulation bayésienne." Grenoble INPG, 2008. https://tel.archives-ouvertes.fr/tel-00361317.
Full textAccurate magnetic resonance brain scan segmentation is critical in a number of clinical and neuroscience applications. This task is challenging due to artifacts, low contrast between tissues and inter-individual variability that inhibit the introduction of a priori knowledge. In this thesis, we propose a new MR brain scan segmentation approach. Unique features of this approach include (1) the coupling of tissue segmentation, structure segmentation and prior knowledge construction, and (2) the consideration of local image properties. Locality is modeled through a multi-agent framework: agents are distributed into the volume and perform a local Markovian segmentation. As an initial approach (LOCUS, Local Cooperative Unified Segmentation), intuitive cooperation and coupling mechanisms are proposed to ensure the consistency of local models. Structures are segmented via the introduction of spatial localization constraints based on fuzzy spatial relations between structures. In a second approach, (LOCUS-B, LOCUS in a Bayesian framework) we consider the introduction of a statistical atlas to describe structures. The problem is reformulated in a Bayesian framework, allowing a statistical formalization of coupling and cooperation. Tissue segmentation, local model regularization, structure segmentation and local affine atlas registration are then coupled in an EM framework and mutually improve. The evaluation on simulated and real images shows good results, and in particular, a robustness to non-uniformity and noise with low computational cost. Local distributed and cooperative MRF models then appear as a powerful and promising approach for medical image segmentation
Pons, Isabelle. "Méthodes de segmentation bayésienne appliquées aux images SAR : théorie et mise en oeuvre." Nice, 1994. http://www.theses.fr/1994NICE4714.
Full textGarcía, Lorenzo Daniel. "Robust segmentation of focal lesions on multi-sequence MRI in multiple sclerosis." Rennes 1, 2010. http://www.theses.fr/2010REN1S018.
Full textMultiple sclerosis (MS) affects around 80. 000 people in France. Magnetic resonance imaging (MRI) is an essential tool for diagnosis of MS and MRI-derived surrogate markers such as MS lesion volumes are often used as measures in MS clinical trials for the development of new treatments. The manual segmentation of these MS lesions is a time-consuming task that shows high inter- and intra-rater variability. We developed an automatic workflow for the segmentation of focal MS lesions on MRI. The segmentation method is based on the robust estimation of a parametric model of the intensities of the brain; lesions are detected as outliers to the model. We proposed two methods to include spatial information in the segmentation using mean shift and graph cut. We performed a quantitative evaluation of our workflow using synthetic and clinical images of two different centers to verify its accuracy and robustness
Garcia, Arnaud. "Analyse statistique et morphologique des images multivaluées : développements logiciels pour les applications cliniques." Phd thesis, École normale supérieure de Cachan - ENS Cachan, 2008. http://tel.archives-ouvertes.fr/tel-00422589.
Full textRémi, Céline. "Contribution au choix de primitives en graphomotricité." Rouen, 1999. http://www.theses.fr/1999ROUES074.
Full textSebbar, Abdeljalil. "Synthèse et segmentation markovienne d'images sur la base d'informations propres à la texture." Compiègne, 1989. http://www.theses.fr/1989COMPD177.
Full textNew algorithms for texture synthesis and segmentation of textured images and for segmentation of noisy images are introduced. Markov random fields are used for texture modelling in a Bayesian approach. For texture synthesis, a set of parameters containing adapted filters and the variances of the corresponding filtered images is computed in a learning step. These parameters are then used in a Markovian model for texture synthesis. So we were able to observe the efficiency of the model and the parameters we choose. Then we have used them in a Bayesian segmentation algorithm of textured images. The originality of these synthesis and segmentation algorithm is that the estimation step is not depending on the number of the grey levels of the texture. The application of the presented synthesis algorithm is limited to homogenous textures with 2 to 4 grey levels, but the segmentation algorithm was applied to homogenous textures with 256 grey levels. An other algorithm for segmentation of noisy images was introduced. The model used is also Markovian. Good results were obtained on a real image
Ribal, Christophe. "Anisotropic neighborhoods of superpixels for thin structure segmentation." Electronic Thesis or Diss., université Paris-Saclay, 2021. http://www.theses.fr/2021UPASG117.
Full textIn the field of computer vision, image segmentation aims at decomposing an image into homogeneous regions. While usually an image is composed of a regular lattice of pixels, this manuscript proposes through the term of site a generic approach able to consider either pixels or superpixels. Robustness to noise in this challenging inverse problem is achieved by formulating the labels as a Markov Random Field, and finding an optimal segmentation under the prior that labels should be homogeneous inside the neighborhood of a site. However, this regularization of the solution introduces unwanted artifacts, such as the early loss of thin structures, defined as structures whose size is small in at least one dimension. Anisotropic neighborhood construction fitted to thin structures allows us to tackle the mentioned artifacts. Firstly, the orientations of the structures in the image are estimated from any of the three presented options: The minimization of an energy, Tensor Voting, and RORPO. Secondly, four methods for constructing the actual neighborhood from the orientation maps are proposed: Shape-based neighborhood, computed from the relative positioning of the sites, dictionary-based neighborhood, derived from the discretization to a finite number of configurations of neighbors for each site, and two path-based neighborhoods, namely target-based neighborhood with fixed extremities, and cardinal-based neighborhood with fixed path lengths. Finally, the results provided by the Maximum A Posteriori criterion (computed with graph cuts optimization) with these anisotropic neighborhoods are compared against isotropic ones on two applications: Thin structure detection and depth reconstruction in Shape From Focus. The different combinations of guidance map estimations and neighborhood constructions are illustrated and evaluated quantitatively and qualitatively in order to exhibit the benefits of the proposed approaches
Văcar, Cornelia Paula. "Inversion for textured images : unsupervised myopic deconvolution, model selection, deconvolution-segmentation." Thesis, Bordeaux, 2014. http://www.theses.fr/2014BORD0131/document.
Full textThis thesis is addressing a series of inverse problems of major importance in the fieldof image processing (image segmentation, model choice, parameter estimation, deconvolution)in the context of textured images. In all of the aforementioned problems theobservations are indirect, i.e., the textured images are affected by a blur and by noise. Thecontributions of this work belong to three main classes: modeling, methodological andalgorithmic. From the modeling standpoint, the contribution consists in the development of a newnon-Gaussian model for textures. The Fourier coefficients of the textured images are modeledby a Scale Mixture of Gaussians Random Field. The Power Spectral Density of thetexture has a parametric form, driven by a set of parameters that encode the texture characteristics.The methodological contribution is threefold and consists in solving three image processingproblems that have not been tackled so far in the context of indirect observationsof textured images. All the proposed methods are Bayesian and are based on the exploitingthe information encoded in the a posteriori law. The first method that is proposed is devotedto the myopic deconvolution of a textured image and the estimation of its parameters.The second method achieves joint model selection and model parameters estimation froman indirect observation of a textured image. Finally, the third method addresses the problemof joint deconvolution and segmentation of an image composed of several texturedregions, while estimating at the same time the parameters of each constituent texture.Last, but not least, the algorithmic contribution is represented by the development ofa new efficient version of the Metropolis Hastings algorithm, with a directional componentof the proposal function based on the”Newton direction” and the Fisher informationmatrix. This particular directional component allows for an efficient exploration of theparameter space and, consequently, increases the convergence speed of the algorithm.To summarize, this work presents a series of methods to solve three image processingproblems in the context of blurry and noisy textured images. Moreover, we present twoconnected contributions, one regarding the texture models andone meant to enhance theperformances of the samplers employed for all of the three methods
Hadrich, Ben Arab Atizez. "Étude des fonctions B-splines pour la fusion d'images segmentées par approche bayésienne." Thesis, Littoral, 2015. http://www.theses.fr/2015DUNK0385/document.
Full textIn this thesis we are treated the problem of nonparametric estimation probability distributions. At first, we assumed that the unknown density f was approximated by a basic mixture quadratic B-spline. Then, we proposed a new estimate of the unknown density function f based on quadratic B-splines, with two methods estimation. The first is based on the maximum likelihood method and the second is based on the Bayesian MAP estimation method. Then we have generalized our estimation study as part of the mixture and we have proposed a new estimator mixture of unknown distributions based on the adapted estimation of two methods. In a second time, we treated the problem of semi supervised statistical segmentation of images based on the hidden Markov model and the B-sline functions. We have shown the contribution of hybridization of the hidden Markov model and B-spline functions in unsupervised Bayesian statistical image segmentation. Thirdly, we presented a fusion approach based on the maximum likelihood method, through the nonparametric estimation of probabilities, for each pixel of the image. We then applied this approach to multi-spectral and multi-temporal images segmented by our nonparametric and unsupervised algorithm
Sodjo, Jessica. "Modèle bayésien non paramétrique pour la segmentation jointe d'un ensemble d'images avec des classes partagées." Thesis, Bordeaux, 2018. http://www.theses.fr/2018BORD0152/document.
Full textThis work concerns the joint segmentation of a set images in a Bayesian framework. The proposed model combines the hierarchical Dirichlet process (HDP) and the Potts random field. Hence, for a set of images, each is divided into homogeneous regions and similar regions between images are grouped into classes. On the one hand, thanks to the HDP, it is not necessary to define a priori the number of regions per image and the number of classes, common or not.On the other hand, the Potts field ensures a spatial consistency. The arising a priori and a posteriori distributions are complex and makes it impossible to compute analytically estimators. A Gibbs algorithm is then proposed to generate samples of the distribution a posteriori. Moreover,a generalized Swendsen-Wang algorithm is developed for a better exploration of the a posteriori distribution. Finally, a sequential Monte Carlo sampler is defined for the estimation of the hyperparameters of the model.These methods have been evaluated on toy examples and natural images. The choice of the best partition is done by minimization of a numbering free criterion. The performance are assessed by metrics well-known in statistics but unused in image segmentation
Kara-Falah, Riad. "Segmentation d'images : coopération, fusion, évaluation." Chambéry, 1995. http://www.theses.fr/1995CHAMS010.
Full textThivin, Solenne. "Détection automatique de cibles dans des fonds complexes. Pour des images ou séquences d'images." Thesis, Université Paris-Saclay (ComUE), 2015. http://www.theses.fr/2015SACLS235/document.
Full textDuring this PHD, we developped an detection algorithm. Our principal objective was to detect small targets in a complex background like clouds for example.For this, we used the spatial covariate structure of the real images.First, we developped a collection of models for this covariate structure. Then, we selected a special model in the previous collection. Once the model selected, we applied the likelihood ratio test to detect the potential targets.We finally studied the performances of our algorithm by testing it on simulated and real images
Zouagui, Tarik. "Approche fonctionnelle générique des méthodes de segmentation d'images." Lyon, INSA, 2004. http://theses.insa-lyon.fr/publication/2004ISAL0042/these.pdf.
Full textImage segmentation is a low-level image processing operation, which consists in recognizing homogeneous regions within an image as distinct and belonging to different objects. A wide range of works has been undertaken to achieve this aim and segmentation has been used in applications ranging from industrial to medical uses. One of the results, is a very great number of segmentation methods, which makes the task of comparing them a very difficult challenge. We propose a new approach of the image segmentation methods based on a functional model (FM). The core of the functional model is a segmentation operator (SO) composed of five elementary blocks called in an iterative process. The functional model unifies segmentation methods under the same framework and allows a better understanding of these methods. Indeed, the decomposition with the same logical way of various segmentation techniques has been obtained and implemented. This showed the genericity of the model and its usefulness in structuring and implementing segmentation methods. We propose also a multi-operator model which represents complex segmentation methods like multiresolution or agent-based methods. The decompositions led to independent functional blocks which have been used to realize a modular software called GenSeg. This software can help in implementing segmentation techniques and in building new methods as well
Precioso, Frédéric. "Contours actifs paramétriques pour la segmentation d'images et vidéos." Nice, 2004. http://www.theses.fr/2004NICE4078.
Full textActive contour modelling represents the main framework of this thesis. Active contours are dynamic methods applied to segmentation of till images and video. The goal is to extract regions corresponding to semantic objects. Image and video segmentation can be cast in a minimization framework by choosing a criterion which includes region and boundary functional. The minimization is achieved through the propagation of a region-based active contour. The efficiency of these methods lies in their robustness and their accuracy. The aim of this thesis is triple : to develop (i) a model of parametric curve providing a smooth active contour, to precise (ii) conditions of stable evolution for such curves, and to reduce (iii) the computation cost of our algorithm in order to provide an efficient solution for real time applications. We mainly consider constraints on contour regularity providing a better robustness regarding to noisy data. In the framework of active contour, we focus on stability of the propagation force, on handling topology changes and convergence conditions. We chose cubic splines curves. Such curves provide great properties of regularity allow an exact computation for analytic expressions involved in the functional and reduce highly the coputation cost. Furthermore, we extended the well-known model-based on interpolating splines to an approximating model based smoothing splines. This latter converts the interpolation error into increased smoothness, smaller energy of the second derivative. The flexibility of this new model provides a tunable balance between accuracy and robustness. The efficiency of implementating such parametric active contour spline-based models has been illustrated for several applications of segmentation process
Koepfler, Georges. "Formalisation et analyse numérique de la segmentation d'images." Paris 9, 1991. https://portail.bu.dauphine.fr/fileviewer/index.php?doc=1991PA090027.
Full textPerroton, Laurent. "Segmentation parallèle d'images volumiques." Lyon 1, 1994. http://www.theses.fr/1994LYO10328.
Full textMoghrani, Madjid. "Segmentation coopérative et adaptative d’images multicomposantes : application aux images CASI." Rennes 1, 2007. http://www.theses.fr/2007REN1S156.
Full textThis thesis focuses on cooperative approaches in image segmentation. Two adaptive systems were implemented; the first is parallel and second is sequential. The parallel system is based on competing methods of segmentation by classification. The sequential system runs these methods according to a predefined schedule. The extraction of features for segmentation is performed according of the region’s nature (uniform or textured). Both systems are composed of three main modules. The first module aims to detect the region’s nature of the image (uniforms or textured) in order to adapt further processings. The second module is dedicated to the segmentation of detected regions according to their nature. The segmentation results are assessed and validated at different levels of the segmentation process. The third module merges intermediate results obtained on the two types of areas. Both systems are tested and compared on synthetic and real mono- and multi-component images issued from aerial remote sensing
Grenier, Thomas. "Apport de l'espace des caractéristiques et des paramètres d'échelle adaptatifs pour le filtrage et la segmentation d'image." Lyon, INSA, 2005. http://theses.insa-lyon.fr/publication/2005ISAL0116/these.pdf.
Full textIn this work, we propose the use of feature space and adaptive scale parameters for image filtering and image segmentation. Feature space is a multidimensional space where a vector xi of parameters related to an entity i can be represented. These parameters, so-called features, represent any information about the entity i such as spatial localisation, colour, texture, curvature, echogenic properties … The scale parameter is a scalar or a matrix quantity used to weight the measurement tied to a given feature. When associated to a particular entity i, the scale parameter becomes an adaptive scale parameter. In a first part, we remind a rigorous definition of the feature space and adaptive scale parameters concepts in statistic, in the context of non-parametric kernel-based estimation theory. Monodimensional, multidimensional and adaptive kernel-based estimations are successively studied. Optimal kernel and optimal scale parameter are defined relatively to a statistic criterion. Then, we present the “Mean Shift” method. Derived from the previous non-parametric estimation, “Mean Shift” method is well-suited to take into account for an entity i both all its features and the locally adaptive behaviour of a scale parameter. “Mean Shift” procedure consists in an iterative mode seeking (local maxima) of the density function. Many “Mean Shift” procedures are detailed and their interests for image filtering presented. In a third part, we propose a formulation to integrate region growing in the feature space and to take into account global or local scale parameters. We also show that a large number of existing region growing methods can be described using this formulation. In the last part, we propose two filtering methods and two region growing methods based on multidimensional and adaptive approaches. Our methods are applied on medical images including US and TEP images giving promising results
Boukala, Nabil. "Contribution des modèles statistiques de forme et d'apparence à la segmentation d'images." Saint-Etienne, 2007. http://www.theses.fr/2007STET4006.
Full textWe purpose in this thesis to study two kinds of deformable models for image segmentation: - active contours or snakes, - statistical models, especially the learning-based active shape (ASM) and active appearance models (AAM). We propose to apply these approaches for segmenting high resolution X-ray images of the bassin. Our comparative study reveals the superiority, in terms of precision and robustness of the ASM over the other studied methods. However, our dataset also points out one major limitation shared by both ASM and AAM approaches which is the need for a large training set. Indeed, for their training, these models require a large number of manually annotated image-examples, thus representing a time-consuming phase and restricting the field of application of these methods. The method we propose, more than an hybrid approach, relies on an original training scheme which leads to improvements in terms of precision of localization and robustness to poor initializations. Our approach is also particularly appropriate for small training sets as, given a suitable PDM, the training of the local appearance models we use can be performed on a single image without producing a significant loss of precision in segmentation results. We have performed extensive experiments to demonstrate our algorithm and are providing the results which also shows promise for applications such as object tracking in a video sequence or surface reconstruction from volume data
Arbelaez, Escalante Pablo Andrés. "Une approche métrique pour la segmentation d'images." Paris 9, 2005. https://portail.bu.dauphine.fr/fileviewer/index.php?doc=2005PA090051.
Full textMonga, Olivier. "Segmentation d'images par croissance hiérarchique de régions." Paris 11, 1988. http://www.theses.fr/1988PA112054.
Full textChabrier, Sébastien. "Contribution à l'évaluation de performances en segmentation d'images." Orléans, 2005. http://www.theses.fr/2005ORLE2076.
Full textPattisina, Ronny. "Codage et segmentation d'images par des méthodes combinatoires." Dijon, 2002. http://www.theses.fr/2002DIJOS053.
Full textDouchez, Marie-Claire. "Utilisation des concepts d'analyse statistique des données et de connexité pour la segmentation des images." Lille 1, 1993. http://www.theses.fr/1993LIL10058.
Full textMartin, Pascal. "Application du principe de minimisation de la complexité stochastique à la segmentation d'images bruitées par contour actif." Aix-Marseille 3, 2006. http://www.theses.fr/2006AIX30010.
Full textImage segmentation consists in divise an image into differents regions of interest. It occurs in many application areas and the processed images can thus be corrupted with noise of various physical origin. Most of the developped segmentation techniques are based on the optimization of a criterion that has at least one parameter to be tune by the user. In this work, we present segmentation algorithms in two regions based on the minimization of the stochastic complexity of the image. In particular, we propose an original nonparametric statistical modelization of the fluctuations of the gray levels. We thus obtain the first segmentation technique adapted to the noise present in the segmented image without \emph{a priori} knowledge of the probability laws which describe it and which is based on the optimization of a criterion without parameter to be tuned by the user
Drot, Sébastien. "Segmentation d'images d'observation de la terre par des techniques de géométrie stochastique." Nice, 2002. http://www.theses.fr/2002NICE5793.
Full textProbabilist techniques used for image segmentation or image classification are generally based on a pixelwise approach. They are well-known for their robustness w. R. T. Noise because of their ability to take into account noise statistics and a priori information on the segmentation (homogeneity, texture). These methods have been proved to be particularly relevant for low and medium resolution data (AVHRR, Landsat, SPOT). With high resolution data (as for example aerial images), these methods reach their limits because they do not use any geometrical information on the underlying objects (in rural environment for example, fields are geometrical entities). So, it is desirable to develop new methods that take into account this kind of information. To keep the advantages of Bayesian approaches within taking into account simple geometrical properties, we suggest to use object point processes. We have developped two families of models based on some equilateral triangles: in the first case, triangles can have only four directions; in the second case, orientation is arbitrary. Those models embed some a priori information that favour configurations closed to a partition of the image (here, a configuration is a collection of triangles); data attachment gives a more important probability for objects that are localized on homogeneous texture. They are optimized thanks to a simulated annealing scheme based on a Reversible Jump MCMC algorithm. At this stage, we obtain an over-segmentation. Then, a post-processing enables to merge areas with similar radiometries. This approach has been applied with success to optical aerial data and radar satellite images
Harp, Josselin. "Contribution à la segmentation des images : applications à l'estimation des plans texturés dans des images planes et omnidirectionnelles." Amiens, 2003. http://www.theses.fr/2003AMIE0313.
Full textEzziane, Nadia. "Segmentation d'images texturées par analyse multi-échelle." Poitiers, 1997. http://www.theses.fr/1997POIT2366.
Full textThourel, Pierre. "Segmentation d'images sonar par modélisation markovienne hiérarchique et analyse multirésolution." Brest, 1996. http://www.theses.fr/1996BRES2026.
Full textZemirli, Kafia. "Contrôle de segmentation pour l'extraction de formes." Paris 11, 2000. http://www.theses.fr/2000PA112292.
Full textGiordana, Nathalie. "Segmentation non supervisee d'images multi-spectrales par chaines de markov cachees." Compiègne, 1996. http://www.theses.fr/1996COMP981S.
Full textTaton, Benjamin. "Modèle déformable à densité adaptative : application à la segmentation d'images." Bordeaux 1, 2004. http://www.theses.fr/2004BOR12872.
Full textLoum, Georges L. "Segmentation pyramidale de textures par décomposition en ondelettes." Paris 12, 1996. http://www.theses.fr/1996PA120021.
Full textVenegas, Martinez Santiago. "Analyse et segmentation de séquences d'images en vue d'une reconnaissance de formes efficace." Paris 5, 2002. http://www.theses.fr/2002PA05S002.
Full textThis work presents a computational technique for tracking moving interfaces. For that, a vecursive linear convolving method to performe arisotropic diffusion in images is presented. The novel approach is that there is not need to estimate local and global properties previously of the concerned evolving interface. The method works on a fixed grid, usually the image pixels grid, and automatically handles changes in the topology of the evolving interface
Dydenko, Igor. "Segmentation dynamique en échocardiographie ultrasonore radiofréquence." Lyon, INSA, 2003. http://theses.insa-lyon.fr/publication/2003ISAL0054/these.pdf.
Full textThe goal of this Ph. D. Thesis is the development of techniques of radiofrequency (RF) image segmentation in cardiac echography. The first part of this work is dedicated to the detection of ultraound contrast agent. A parametric method based on local spectral autoregressive (AR) analysis of the RF signal is proposed. It is shown on simulations and in vitro images thats the proposed approach is stable with respect to concentration of the agent and the instrumental MI. The second part of the work concerns segmentation of cardiac RF sequences. Based on AR spectral analysis, it is shown that the spectral contents of the RF signal brings complementary information, as compared to the envelope image alone. A segmentation method is subsequently introduced, based on the level set framework coupled with affine registration. The method is validated on numerical simulations as well as on ultrasound in vivo sequences, showing its interest for segmentation and tracking of the cardiac muscle
Masson, Pascale. "Etude d'algorithmes de classification contextuelle et application à la segmentation d'images satellite." Brest, 1991. http://www.theses.fr/1991BRES2007.
Full textVandenbroucke, Nicolas. "Segmentation d'images couleur par classification de pixels dans des espaces d'attributs colorimétriques adaptés : application à l'analyse d'images de football." Lille 1, 2000. https://pepite-depot.univ-lille.fr/LIBRE/Th_Num/2000/50376-2000-404.pdf.
Full textThieu, Quang Tung. "Segmentation by convex active contour models : application to skin lesion and medical images." Paris 13, 2013. http://www.theses.fr/2013PA132063.
Full textGaucher, Pierre. "Segmentation d'images numeriques par mesure de proximite de deux pixels adjacents." Tours, 1992. http://www.theses.fr/1992TOUR4002.
Full textCoquin, Didier. "Segmentation et analyse d'Images pour la classification automatique : application au zooplancton." Rennes 1, 1991. http://www.theses.fr/1991REN10090.
Full textM'Hiri, Slim. "Segmentation d'images par classification floue fondée sur une approche neuromimétique." Paris 12, 1996. http://www.theses.fr/1996PA120082.
Full textLu, Jun-Wei. "Segmentation d'images couleur et application à la séparation des oignons et des adventices." Dijon, 2003. http://www.theses.fr/2003DIJOS003.
Full textColantoni, Philippe. "Contribution des structures de données à la segmentation d'images couleur : élaboration d'un outil d'infographie textile." Saint-Etienne, 1998. http://www.theses.fr/1998STET4013.
Full textFontaine, Michaël. "Segmentation non supervisée d'images couleur par analyse de la connexité des pixels." Lille 1, 2001. https://pepite-depot.univ-lille.fr/LIBRE/Th_Num/2001/50376-2001-305-306.pdf.
Full text