Dissertationen zum Thema „Algorithme de suivi d'objets“
Geben Sie eine Quelle nach APA, MLA, Chicago, Harvard und anderen Zitierweisen an
Machen Sie sich mit Top-50 Dissertationen für die Forschung zum Thema "Algorithme de suivi d'objets" bekannt.
Neben jedem Werk im Literaturverzeichnis ist die Option "Zur Bibliographie hinzufügen" verfügbar. Nutzen Sie sie, wird Ihre bibliographische Angabe des gewählten Werkes nach der nötigen Zitierweise (APA, MLA, Harvard, Chicago, Vancouver usw.) automatisch gestaltet.
Sie können auch den vollen Text der wissenschaftlichen Publikation im PDF-Format herunterladen und eine Online-Annotation der Arbeit lesen, wenn die relevanten Parameter in den Metadaten verfügbar sind.
Sehen Sie die Dissertationen für verschiedene Spezialgebieten durch und erstellen Sie Ihre Bibliographie auf korrekte Weise.
Delamarre, Quentin. „Suivi du mouvement d'objets articulés dans des séquences d'images vidéo“. Nice, 2003. http://www.theses.fr/2003NICE4067.
Der volle Inhalt der QuelleWe introduce you to a method designed to provide the computer, the ability to automatically understand the motions of a markerless filmed human in a multi-cameras environment. In a first step, we decide to estimate the time extended motions variations. Secondly, these variations are interpreted in order to give them significance. After being retrieved from the video pictures, information are compared to a tracked object geometric model. These information could be a segmentation of the object silhouette in case of far distant cameras from each other, or a three-dimensional reconstruction in case of near distant cameras from each other. We ll explain why this distinction is made. In each case, we assume that the object geometry is known thanks to the build of a 3D model made of simple articulated rigid parts. The position parameters error correction is done by creating forces and by resolving the equations of the 3D articulated model dynamic. Information found in the pictures allow us to create such forces. Different steps of the algorithm are introduced: how to calibrate the cameras, the 3D model structure, the information retrieve process, how to apply forces to the model, its dynamic, the estimation and anticipation of the motion in the scene thanks to a Kalman filter. Finally, we expose encouraging results and try to give ideas in order to generalize the algorithm
Loesch, Angélique. „Localisation d'objets 3D industriels à l'aide d'un algorithme de SLAM contraint au modèle“. Thesis, Université Clermont Auvergne (2017-2020), 2017. http://www.theses.fr/2017CLFAC059/document.
Der volle Inhalt der QuelleIn the industry domain, applications such as quality control, automation of complex tasks or maintenance support with Augmented Reality (AR) could greatly benefit from visual tracking of 3D objects. However, this technology is under-exploited due to the difficulty of providing deployment easiness, localization quality and genericity simultaneously. Most existing solutions indeed involve a complex or an expensive deployment of motion capture sensors, or require human supervision to simplify the 3D model. And finally, most tracking solutions are restricted to textured or polyhedral objects to achieved an accurate camera pose estimation.Tracking any object is a challenging task due to the large variety of object forms and appearances. Industrial objects may indeed have sharp edges, or occluding contours that correspond to non-static and view-point dependent edges. They may also be textured or textureless. Moreover, some applications require to take large amplitude motions as well as object occlusions into account, tasks that are not always dealt with common model-based tracking methods. These approaches indeed exploit 3D features extracted from a model, that are matched with 2D features in the image of a video-stream. However the accuracy and robustness of the camera localization depend on the visibility of the object as well as on the motion of the camera. To better constrain the localization when the object is static, recent solutions rely on environment features that are reconstructed online, in addition to the model ones. These approaches combine SLAM (Simultaneous Localization And Mapping) and model-based tracking solutions by using constraints from the 3D model of the object of interest. Constraining SLAM algorithms with a 3D model results in a drift free localization. However, such approaches are not generic since they are only adapted for textured or polyhedral objects. Furthermore, using the 3D model to constrain the optimization process may generate high memory consumption,and limit the optimization to a temporal window of few cameras. In this thesis, we propose a solution that fulfills the requirements concerning deployment easiness, localization quality and genericity. This solution, based on a visual key-frame-based constrained SLAM, only exploits an RGB camera and a geometric CAD model of the static object of interest. An RGB camera is indeed preferred over an RGBD sensor, since the latter imposes limits on the volume, the reflectiveness or the absorptiveness of the object, and the lighting conditions. A geometric CAD model is also preferred over a textured model since textures may hardly be considered as stable in time (deterioration, marks,...) and may vary for one manufactured object. Furthermore, textured CAD models are currently not widely spread. Contrarily to previous methods, the presented approach deals with polyhedral and curved objects by extracting dynamically 3D contour points from a model rendered on GPU. This extraction is integrated as a structure constraint into the constrained bundle adjustment of a SLAM algorithm. Moreover we propose different formalisms of this constraint to reduce the memory consumption of the optimization process. These formalisms correspond to hybrid structure/trajectory constraints, that uses output camera poses of a model-based tracker. These formalisms take into account the structure information given by the 3D model while relying on the formalism of trajectory constraints. The proposed solution is real-time, accurate and robust to occlusion or sudden motion. It has been evaluated on synthetic and real sequences of different kind of objects. The results show that the accuracy achieved on the camera trajectory is sufficient to ensure a solution perfectly adapted for high-quality Augmented Reality experiences for the industry
Lafaye, de Micheaux Hugo. „Traitement d'images pour la ségrégation en transport de sédiments par charriage : morphologie et suivi d'objets“. Thesis, Lyon, 2017. http://www.theses.fr/2017LYSES008/document.
Der volle Inhalt der QuelleSediment transport in rivers and mountain streams remains poorly understood partly due to the polydispersity of particles and resulting segregation. Experiments in a channel were carried out to study bedload transport of bimodal bead mixtures. The behavior of the beads is recorded through video sequences. This work is about the development of image processing methods to analyse the obtained data. Firstly, we developed a method of image segmentation to study the infiltration of fine particles and its influence on the evolution of bed mobility. Thanks to this method, an experimental study shows that the bed slope evolution follows an exponential decay. Secondly, we optimised deterministic tracking algorithms to enable the study of trajectories on long-duration phenomena of segregation, which was not possible with previous work done at Irstea. Moreover we set up relevant evaluation measures and elaborated ground truth sequences to quantify the results. We observed benefits in execution time, consistency, precision and memory. Thirdly, we developed a new algorithm based on multiple model particle filtering to better deal with complex dynamics of particles and to gain robustness. This approach allows taking unreliable detections into account, correcting them and thus avoiding difficulties in the target tracking as encountered with the deterministic algorithm
Nghiem, Anh-Tuan. „Algorithmes adaptatifs d'estimation du fond pour la détection des objets mobiles dans les séquences vidéos“. Phd thesis, Université de Nice Sophia-Antipolis, 2010. http://tel.archives-ouvertes.fr/tel-00505881.
Der volle Inhalt der QuelleDulac, William. „Méthodes pour l'évaluation de l'activité cyclonique tropicale en changement climatique“. Electronic Thesis or Diss., Toulouse 3, 2023. http://www.theses.fr/2023TOU30315.
Der volle Inhalt der QuelleGiven their devastating impact on the populations and infrastructures of the countries concerned the future evolution of tropical cyclone activity in the context of global warming is an issue of great importance. Two methods exist for assessing tropical cyclone activity under climate change in climate models: the use of cyclone detection algorithms (trackers) or the use of cyclogenesis indices, which translate statistical relationships linking observed cyclone activity to large-scale atmospheric variables. These two methods tend to provide opposite projections in climate simulations. Motivated by this disagreement, this thesis proposes to explore these two approaches, with the aim of making improvements to each. Firstly, the CNRM tropical cyclone tracker is applied to the ERA5 reanalysis of the European Centre for Medium-Range Weather Forecasts, and evaluated using the IBTrACS database of cyclone observations. Its performance is evaluated in terms of detection probability and false alarm rate (POD and FAR), after optimizing detection parameters and applying an appropriate mid-latitude system filter. Several metrics for assessing the similarity of the tracks detected in ERA5 with those observed are then proposed and compared. These innovative metrics are complementary to POD and FAR, and show that optimizing detection parameters is accompanied by a slight improvement in track similarity. New cyclogenesis indices are then constructed on ERA5 by Poisson regression between large-scale thermal and dynamic predictors, and the IBTrACS database. The regressions are run at different spatial and temporal resolutions, as well as on a global scale and for different ocean basins. The increased temporal resolution enables the equatorial bias present in the most commonly used indices to be corrected. However, the interannual variability of the indices appears to be robust to changes in the weighting coefficients of the large-scale variables. Following this observation, the contribution of adding predictors to the regressions is evaluated on ERA5 as well as in the ARPEGE model; on the one hand by explicitly adding a diagnostic of the El Niño (ENSO) variability mode to the index, and on the other hand by replacing the relative humidity at 600 hPa by the integrated moisture saturation deficit on the column (VPD). The addition of ENSO diagnostics improves the interannual variability of the index in most ocean basins. Correlations with observed series are made statistically significant at the 95% threshold in all basins except the North Atlantic. The use of the VPD cancels out the upward trends in the historical period observed in indices based on relative humidity. The resulting index is therefore in better agreement with observations. When applied to very high-resolution ARPEGE climate simulations, under the RCP8.5 scenario, the VPD also amplifies the decrease in cyclonic activity
Reverter, Valeiras David. „Event-based detection and tracking“. Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066566/document.
Der volle Inhalt der QuelleThe main goal of this thesis is the development of event-based algorithms for visual detection and tracking. This algorithms are specifically designed to work on the output of neuromorphic event-based cameras. This type of cameras are a new type of bioinspired sensors, whose principle of operation is based on the functioning of the retina: every pixel is independent and generates events asynchronously when a sufficient amount of change is detected in the luminance at the corresponding position on the focal plane. This new way of encoding visual information calls for new processing methods. First, a part-based shape tracking is presented, which represents an object as a set of simple shapes linked by springs. The resulting virtual mechanical system is simulated with every incoming event. Next, a line and segment detection algorithm is introduced, which can be employed as an event-based low level feature. Two event-based methods for 3D pose estimation are then presented. The first of these 3D algorithms is based on the assumption that the current estimation is close to the true pose of the object, and it consequently requires a manual initialization step. The second of the 3D methods is designed to overcome this limitation. All the presented methods update the estimated position (2D or 3D) of the tracked object with every incoming event. This results in a series of trackers capable of estimating the position of the tracked object with microsecond resolution. This thesis shows that event-based vision allows to reformulate a broad set of computer vision problems, often resulting in simpler but accurate algorithms
Benahmed, Daho Omar. „Radar ULB pour la vision à travers les murs : mise au point d'une chaîne de traitement de l'information d'un radar imageur“. Thesis, La Rochelle, 2014. http://www.theses.fr/2014LAROS036/document.
Der volle Inhalt der QuelleThis report is focused on Through-the-wall surveillance (TTS) using UWB radar, with the objective of developing a complete information processing pipeline (IPP) which can be used by different types of imaging radar. To do this, we want to take into account any a priori information, nor on the target, or their environmental context. In addition, the IPP must meet criteria of adaptability and modularity to process information from two types of radar, including pulsed and FMCW developed in two projects that are part of the work of this thesis. Radar imaging is an important point in this context ; we approach it by combining backprojection and trilateration algorithms and show the improvement with the use of a CFAR detector taking into account the shape of the targets signatures.The development of the IPP is our main contribution. The flow of radar images obtained is divided into two parts. The first dynamic sequence contains moving targets are tracked by a multiple hypothesis approach. The second static sequence contains stationary targets and interior walls that are highlighted by Radon transformbases approach. We developed a simulator operating in time and frequency domain to design the algorithms of the IPP and test their robustness. Several simulated scenarios and experimental measurements show that our IPP is relevant and robust. It is thus validated for both radar systems
Jacquot, Aude. „Suivi d'objets en imagerie aérienne“. Phd thesis, Grenoble INPG, 2006. http://tel.archives-ouvertes.fr/tel-00379479.
Der volle Inhalt der QuelleLe thème principal de cette thèse est donc le suivi d'objets à partir d'images aériennes. Nous souhaitons utiliser la faisabilité d'une extraction d'information 3D à partir de séquences vidéo afin d'améliorer les algorithmes de suivi de matériels aéroportés existants. Pour cela, nous nous plaçons dans un cadre bayésien et formulons le suivi de manière probabiliste, au moyen d'un filtre particulaire. Nous avons mis en place trois algorithmes:
Le premier est basé sur des histogrammes de couleurs, que l'on combine à un filtrage particulaire;
c'est un suivi purement 2D dans le sens où aucune information 3D réelle de la scène est utilisée.
Le second est basé sur des modèles géométriques (qui peuvent être 2D ou 3D), que l'on combine à un filtrage particulaire. Nous ajoutons une étape supplémentaire au filtrage particulaire classique, nous permettant de changer de modèle lorsque l'algorithme le juge nécessaire.
Enfin, le dernier algorithme combine les deux précédents; l'intégration d'histogrammes de couleurs et d'informations de contours dans un filtre particulaire permet non seulement de rendre le suivi d'objets plus robuste, mais aussi de prendre en compte de l'information 3D réelle de la scène observée.
Un protocole d'évaluation a été mis en place pour évaluer les performances de ces algorithmes. Des résultats illustrent les performances de ces algorithmes.
Romero, Mier y. Teran Andrés. „Real-time multi-target tracking : a study on color-texture covariance matrices and descriptor/operator switching“. Phd thesis, Université Paris Sud - Paris XI, 2013. http://tel.archives-ouvertes.fr/tel-01002065.
Der volle Inhalt der QuelleKumar, Ratnesh. „Segmentation vidéo et suivi d'objets multiples“. Thesis, Nice, 2014. http://www.theses.fr/2014NICE4135/document.
Der volle Inhalt der QuelleIn this thesis we propose novel algorithms for video analysis. The first contribution of this thesis is in the domain of video segmentation wherein the objective is to obtain a dense and coherent spatio-temporal segmentation. We propose joining both spatial and temporal aspects of a video into a single notion Fiber. A fiber is a set of trajectories which are spatially connected by a mesh. Fibers are built by jointly assessing spatial and temporal aspects of the video. Compared to the state-of-the-art, a fiber based video segmentation presents advantages such as a natural spatio-temporal neighborhood accessor by a mesh, and temporal correspondences for most pixels in the video. Furthermore, this fiber-based segmentation is of quasi-linear complexity w.r.t. the number of pixels. The second contribution is in the realm of multiple object tracking. We proposed a tracking approach which utilizes cues from point tracks, kinematics of moving objects and global appearance of detections. Unification of all these cues is performed on a Conditional Random Field. Subsequently this model is optimized by a combination of message passing and an Iterated Conditional Modes (ICM) variant to infer object-trajectories. A third, minor, contribution relates to the development of suitable feature descriptor for appearance matching of persons. All of our proposed approaches achieve competitive and better results (both qualitatively and quantitatively) than state-of-the-art on open source datasets
Villeneuve, Guillaume. „Détection d'objets multi-parties par algorithme adaptatif et optimisé“. Thesis, Université Laval, 2012. http://www.theses.ulaval.ca/2012/29483/29483.pdf.
Der volle Inhalt der QuelleIn this thesis, we propose improvements to an existing unknown shape object detection method that uses simple primitives. Firstly, we eliminate cases where no results were obtained with some images using an adaptive algorithm by removing most of the fixed thresholds, assuring a certain number of primitive groups at each step. Secondly, adding some optimizations and a parallel version of the algorithm make the running time of this new algorithm reasonable. Thirdly, we approach the problem of the redundant solutions by adding a new structuring step that will reduce their number without affecting their variety using hierarchical clustering. Finally, we adjust some parameters and results are produced using three sets of 10 images. We prove in an objective manner that the obtained results are better than those of the previous method.
Masson, Lucie. „Suivi temps-réel d'objets 3D pour la réalité augmentée“. Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2005. http://tel.archives-ouvertes.fr/tel-00685727.
Der volle Inhalt der QuellePINEAU, PATRICK. „Detection et suivi d'objets par analyse de sequences d'images“. Rennes 1, 1991. http://www.theses.fr/1991REN10111.
Der volle Inhalt der QuelleLACASSAGNE, LIONEL. „Detection de mouvement et suivi d'objets en temps reel“. Paris 6, 2000. http://www.theses.fr/2000PA066252.
Der volle Inhalt der QuelleAllezard, Nicolas. „Reconnaissance,localisation et suivi d'objets texturés par vision monoculaire“. Clermont Ferrand 2, 2001. http://www.theses.fr/2001CLF21314.
Der volle Inhalt der QuelleThis thesis deals with textured objects recognition by monocular vision as well as their localization and tracking in the camera coordinates system. We have developed a mixed approach, based on the matching of visual primitives between the image to analyse and the images of a training base, but which uses the knowledge of the 3D object model to validate the result of the recognition process and to locate the object. In a context of textured objects the most relevant features seems to be keypoints. These points are located in image zones having a high information content. We propose a local description of such features directly based on the luminance signal. It is insensitive to image rotations ans translations and we also present a multi-scale implementation of these description in order to recognize objects in spite of important scale changes. Before the recognition, a training stage is carried out. It consists in the genration of synthetic images of the object then in the creation of a data base gathering the extracted primitives and their characterization vector. After that the data are organize in order to scan the base as fast as possible. The use of synthetics images allows them to compute the 3D coordinates of the keypoints extracted at the training stage. So, after the matching done, we obtain a set of 2D/3D pairs between the image and the object model. Thus, we compute the object position with a robust method which introduce a strong geometrical constraint on the 2D/3D pairs. In the case of a images sequence, once the object position known in the first image, the process can be considerably accelerated : the features extraction is limited to a part of the image and the data base scanning is initialized by the preceding 3D position of the object
Laguzet, Florence. „Etude et optimisation d'algorithmes pour le suivi d'objets couleur“. Thesis, Paris 11, 2013. http://www.theses.fr/2013PA112197.
Der volle Inhalt der QuelleThe work of this thesis focuses on the improvement and optimization of the Mean-Shift color object tracking algorithm, both from a theoretical and architectural point of view to improve both the accuracy and the execution speed. The first part of the work consisted in improving the robustness of the tracking. For this, the impact of color space representation on the quality of tracking has been studied, and a method for the selection of the color space that best represents the object to be tracked has been proposed. The method has been coupled with a strategy determining the appropriate time to recalculate the model. Color space selection method was also used in collaboration with another object tracking algorithm to further improve the tracking robustness for particularly difficult sequences : the covariance tracking which is more time consuming. The objective of this work is to obtain an entire real time system running on multi-core SIMD processors. A study and optimization phase has been made in order to obtain algorithms with a complexity that is configurable so that they can run in real time on different platforms, for various sizes of images and object tracking. In this context of compromise between speed and performance, it becomes possible to do real-time tracking on processors like ARM Cortex A9
Gomila, Cristina. „Mise en correspondance de partitions en vue du suivi d'objets“. Phd thesis, École Nationale Supérieure des Mines de Paris, 2001. http://pastel.archives-ouvertes.fr/pastel-00003272.
Der volle Inhalt der QuelleSekkal, Rafiq. „Techniques visuelles pour la détection et le suivi d'objets 2D“. Phd thesis, INSA de Rennes, 2014. http://tel.archives-ouvertes.fr/tel-00981107.
Der volle Inhalt der QuelleDziri, Aziz. „Suivi visuel d'objets dans un réseau de caméras intelligentes embarquées“. Thesis, Clermont-Ferrand 2, 2015. http://www.theses.fr/2015CLF22610/document.
Der volle Inhalt der QuelleMulti-object tracking constitutes a major step in several computer vision applications. The requirements of these applications in terms of performance, processing time, energy consumption and the ease of deployment of a visual tracking system, make the use of low power embedded platforms essential. In this thesis, we designed a multi-object tracking system that achieves real time processing on a low cost and a low power embedded smart camera. The tracking pipeline was extended to work in a network of cameras with nonoverlapping field of views. The tracking pipeline is composed of a detection module based on a background subtraction method and on a tracker using the probabilistic Gaussian Mixture Probability Hypothesis Density (GMPHD) filter. The background subtraction, we developed, is a combination of the segmentation resulted from the Zipfian Sigma-Delta method with the gradient of the input image. This combination allows reliable detection with low computing complexity. The output of the background subtraction is processed using a connected components analysis algorithm to extract the features of moving objects. The features are used as input to an improved version of GMPHD filter. Indeed, the original GMPHD do not manage occlusion problems. We integrated two new modules in GMPHD filter to handle occlusions between objects. If there are no occlusions, the motion feature of objects is used for tracking. When an occlusion is detected, the appearance features of the objects are saved to be used for re-identification at the end of the occlusion. The proposed tracking pipeline was optimized and implemented on an embedded smart camera composed of the Raspberry Pi version 1 board and the camera module RaspiCam. The results show that besides the low complexity of the pipeline, the tracking quality of our method is close to the stat of the art methods. A frame rate of 15 − 30 was achieved on the smart camera depending on the image resolution. In the second part of the thesis, we designed a distributed approach for multi-object tracking in a network of non-overlapping cameras. The approach was developed based on the fact that each camera in the network runs a GMPHD filter as a tracker. Our approach is based on a probabilistic formulation that models the correspondences between objects as an appearance probability and space-time probability. The appearance of an object is represented by a vector of m dimension, which can be considered as a histogram. The space-time features are represented by the transition time between two input-output regions in the network and the transition probability from a region to another. Transition time is modeled as a Gaussian distribution with known mean and covariance. The distributed aspect of the proposed approach allows a tracking over the network with few communications between the cameras. Several simulations were performed to validate the approach. The obtained results are promising for the use of this approach in a real network of smart cameras
Duculty, Florent. „Suivi automatique d'objets 3D basé sur l'apparence dans des séquences d'images“. Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2003. http://tel.archives-ouvertes.fr/tel-00610720.
Der volle Inhalt der QuelleAlmanza-Ojeda, Dora Luz. „Détection et suivi d'objets mobiles perçus depuis un capteur visuel embarqué“. Phd thesis, Université Paul Sabatier - Toulouse III, 2011. http://tel.archives-ouvertes.fr/tel-01017785.
Der volle Inhalt der QuelleAlmanza, Ojeda Dora Luz. „Détection et suivi d'objets mobiles perçus depuis un capteur visuel embarqué“. Toulouse 3, 2011. http://thesesups.ups-tlse.fr/2339/.
Der volle Inhalt der QuelleThis dissertation concerns the detection and the tracking of mobile objets in a dynamic environment, using a camera embedded on a mobile robot. It is an important challenge because only a single camera is used to solve the problem. We must detect mobile objects in the scene, analyzing their apparent motions on images, excluding the motion caused by the ego-motion of the camera. First it is proposed a spatio-remporal analysis of the image sequence based on the sparse optical flow. The a contrario clustering method provides the grouping of dynamic points, without using a priori information and without parameter tuning. This method success is based on the accretion of sufficient information on positions and velocities of these points. We call tracking time, the time required in order to acquire images analyzed to provide the points characterization. A probabilistic map is built in order to find image areas with the higher probabilities to find a mobile objet; this map allows an active selection of new points close the previously detected mobile regions, making larger these regions. In a second step, it is proposed an iterative approach to perform the detection-clustering-tracking process on image sequences acquired from a fixed camera for indoor or outdoor applications. An object is described by an active contour, updated so that the initial object model remains inside the contour. Finally it is presented experimental results obtained on images acquired from a camera embedded on a mobile robot navigating in outdoor environments with rigid or non rigid mobile objects ; it is shown that the method works to detect obstacles during the navigation in a priori unknown environments, first with a weak speed, then with more a realistic speed, compensating the robot ego-motion in images
BRAUD, PASCAL. „Reconnaissance, localisation et suivi d'objets polyedriques modelises par vision multi-oculaire“. Clermont-Ferrand 2, 1996. http://www.theses.fr/1996CLF21787.
Der volle Inhalt der QuelleLlucia, Ludovic. „Suivi d'objets à partir d'images issues de caméras mobiles non calibrées“. Thesis, Aix-Marseille 2, 2011. http://www.theses.fr/2011AIX22009/document.
Der volle Inhalt der QuelleThis work refers to a 3D simulator that has for purpose to help football trainers interpreting tactical sequences based on real situations. This simulator has to be able to interpret video movies, to reconstruct a situation. The camera’s calibration state has to be as simple as possible. The first part of this document refers to the solution elaborated to implement this constraint whereas the second one is more oriented on the industrialisation process. These processes imply to focus on vision computing and ergonomics problems and to answer questions such as : how to characterize a homographic transformation matching the image and the model ? How to retrieve the position of the camera? Which area is part of the image? In an ergonomically point of view, the simulator has to reproduce the game play reality and to improve the abstraction and the communication of the coaches
Bouttefroy, Philippe. „Suivi visuel d'objets et détection de comportements anormaux par inférence contextuelle Bayesienne“. Phd thesis, Université Paris-Nord - Paris XIII, 2010. http://tel.archives-ouvertes.fr/tel-00562299.
Der volle Inhalt der QuelleRogez, Matthieu. „Utilisation du contexte pour la détection et le suivi d'objets en vidéosurveillance“. Thesis, Lyon 2, 2015. http://www.theses.fr/2015LYO22006.
Der volle Inhalt der QuelleVideo-surveillance cameras are increasingly used in our environment. They are indeed present almost everywhere in the cities, supermarkets, airports, warehouses, etc. These cameras are used, among other things, in order to detect suspect behavior (an intrusion for instance) or to recognize a specific category of object or person (gender detection, license plates detection). Other applications also exist to count and/or track people in order to analyze their behavior. Due to the increasing number of cameras and the difficulty to achieve these tasks manually, several video analysis methods have been developed in order to address them automatically. In this thesis, we mainly focus on the detection and tracking of moving objects from a fixed camera. Unlike methods based solely on images captured by cameras, our approach integrates contextual pieces of information in order better interpret these images. Thus we propose to build a geometric and geolocalized model of the scene and the camera. This model is built directly from the pre-deployment studies of the cameras and uses the OpenStreetMap geographical database to build 3d models of buildings near the camera. We added to this model the ability to predict the position of the sun throughout the day and the resulting shadows in the scene. By predicting the shadows, and deleting them from the foreground mask, our method is able to improve the segmentation of pedestrians. Regarding the tracking of multiple mobile objects, we use the formalism of finite state machines to effectively model the states and possible transitions that an object is allowed to take. This allows us to tailor the processing of each object according to its state. We manage the inter-object occlusion using a collective tracking strategy. When taking part in an occlusion, objects are regrouped and tracked collectively. At the end of the occlusion, each object is re-identified and individual tracking resume. Our algorithm adapts to any type of ground-moving object (pedestrians, vehicles, etc.) and seamlessly integrates in the developed scene model. We have also developed several retro-actions taking advantage of the knowledge of tracked objects to improve the detections obtained with the background model. In particular, we tackle the issue of stationary objects often integrated erroneously in the background and we revisited the initial proposal regarding shadow removal. All proposed solutions have been implemented in the Foxstream products and are able to run in real-time
Mikram, Mounia. „Suivi d'objets dans une séquence d'images par modèle d'apparence : conception et évaluation“. Thesis, Bordeaux 1, 2008. http://www.theses.fr/2008BOR13736/document.
Der volle Inhalt der QuelleAbstract
Weiss, Marie. „DEVELOPPEMENT D'UN ALGORITHME DE SUIVI DE LA VEGETATION A LARGE ECHELLE“. Phd thesis, Université de Nice Sophia-Antipolis, 1998. http://tel.archives-ouvertes.fr/tel-00707683.
Der volle Inhalt der QuelleWEISS, MARIE. „Developpement d'un algorithme de suivi de la vegetation a large echelle“. Nice, 1998. http://www.theses.fr/1998NICE5323.
Der volle Inhalt der QuelleVincent, Garcia. „Suivi d'objets d'intérêt dans une séquence d'images : des points saillants aux mesures statistiques“. Phd thesis, Université de Nice Sophia-Antipolis, 2008. http://tel.archives-ouvertes.fr/tel-00374657.
Der volle Inhalt der QuelleLa première méthode repose sur l'analyse de trajectoires temporelles de points saillants et réalise un suivi de régions d'intérêt. Des points saillants (typiquement des lieux de forte courbure des lignes isointensité) sont détectés dans toutes les images de la séquence. Les trajectoires sont construites en liant les points des images successives dont les voisinages sont cohérents. Notre contribution réside premièrement dans l'analyse des trajectoires sur un groupe d'images, ce qui améliore la qualité d'estimation du mouvement. De plus, nous utilisons une pondération spatio-temporelle pour chaque trajectoire qui permet d'ajouter une contrainte temporelle sur le mouvement tout en prenant en compte les déformations géométriques locales de l'objet ignorées par un modèle de mouvement global.
La seconde méthode réalise une segmentation spatio-temporelle. Elle repose sur l'estimation du mouvement du contour de l'objet en s'appuyant sur l'information contenue dans une couronne qui s'étend de part et d'autre de ce contour. Cette couronne nous renseigne sur le contraste entre le fond et l'objet dans un contexte local. C'est là notre première contribution. De plus, la mise en correspondance par une mesure de similarité statistique, à savoir l'entropie du résiduel, d'une portion de la couronne et d'une zone de l'image suivante dans la séquence permet d'améliorer le suivi tout en facilitant le choix de la taille optimale de la couronne.
Enfin, nous proposons une implémentation rapide d'une méthode de suivi de régions d'intérêt existante. Cette méthode repose sur l'utilisation d'une mesure de similarité statistique : la divergence de Kullback-Leibler. Cette divergence peut être estimée dans un espace de haute dimension à l'aide de multiples calculs de distances au k-ème plus proche voisin dans cet espace. Ces calculs étant très coûteux, nous proposons une implémentation parallèle sur GPU (grâce à l'interface logiciel CUDA de NVIDIA) de la recherche exhaustive des k plus proches voisins. Nous montrons que cette implémentation permet d'accélérer le suivi des objets, jusqu'à un facteur 15 par rapport à une implémentation de cette recherche nécessitant au préalable une structuration des données.
Hachour, Samir. „Suivi et classification d'objets multiples : contributions avec la théorie des fonctions de croyance“. Thesis, Artois, 2015. http://www.theses.fr/2015ARTO0206/document.
Der volle Inhalt der QuelleThis thesis deals with multi-objet tracking and classification problem. It was shown that belieffunctions allow the results of classical Bayesian methods to be improved. In particular, a recentapproach dedicated to a single object classification which is extended to multi-object framework. Itwas shown that detected observations to known objects assignment is a fundamental issue in multiobjecttracking and classification solutions. New assignment solutions based on belief functionsare proposed in this thesis, they are shown to be more robust than the other credal solutions fromrecent literature. Finally, the issue of multi-sensor classification that requires a second phase ofassignment is addressed. In the latter case, two different multi-sensor architectures are proposed, aso-called centralized one and another said distributed. Many comparisons illustrate the importanceof this work, in both situations of constant and changing objects classes
Hua, Yang. „Vers un suivi robuste d'objets visuels : sélection de propositions et traitement des occlusions“. Thesis, Université Grenoble Alpes (ComUE), 2016. http://www.theses.fr/2016GREAM012/document.
Der volle Inhalt der QuelleIn this dissertation we address the problem of visual object tracking, whereinthe goal is to localize an object and determine its trajectory over time. Inparticular, we focus on challenging scenarios where the object undergoessignificant transformations, becomes occluded or leaves the field of view. Tothis end, we propose two robust methods which learn a model for the object ofinterest and update it, to reflect its changes over time.Our first method addresses the tracking problem in the context of objectsundergoing severe geometric transformations, such as rotation, change in scale.We present a novel proposal-selection algorithm, which extends the traditionaldiscriminative tracking-by-detection approach. This method proceeds in twostages -- proposal followed by selection. In the proposal stage, we compute acandidate pool that represents the potential locations of the object byrobustly estimating the geometric transformations. The best proposal is thenselected from this candidate set to localize the object precisely usingmultiple appearance and motion cues.Second, we consider the problem of model update in visual tracking, i.e.,determining when to update the model of the target, which may become occludedor leave the field of view. To address this, we use motion cues to identify thestate of the object in a principled way, and update the model only when theobject is fully visible. In particular, we utilize long-term trajectories incombination with a graph-cut based technique to estimate parts of the objectsthat are visible.We have evaluated both our approaches extensively on several trackingbenchmarks, notably, recent online tracking benchmark and the visual objecttracking challenge datasets. Both our approaches compare favorably to thestate of the art and show significant improvement over several other recenttrackers. Specifically, our submission to the visual object tracking challengeorganized in 2015 was the winner in one of the competitions
Pressigout, Muriel. „Approches hybrides pour le suivi temps-réel d'objets complexes dans des séquences vidéos“. Rennes 1, 2006. http://www.theses.fr/2006REN1S106.
Der volle Inhalt der QuelleGarcia, Vincent. „Suivi d'objets d'intérêt dans une séquence d'images : des points saillants aux mesures statistiques“. Nice, 2008. http://www.theses.fr/2008NICE4059.
Der volle Inhalt der QuelleThe problem of object tracking is a problem arising in domains such as computer vision (video surveillance for instance) and cinematographic post-production (special effects). There are two major classes of solution to this problem : region of interest tracking, which indicates a coarse tracking, and space-time segmentation, which corresponds to a precise tracking of the region of interest’s contour. In both cases, the region of interest must be selected beforehand on the first, and possibly on the last image of the video sequence. In this thesis, we propose two tracking methods (one of each type). We propose also a fast implementation of an existing tracking method using the programmation on Graphics Processing Unit (GPU). The first method is based on the analysis of temporal trajectories of salient points and provides a region of interest tracking. Salient points (typically of point of strong curvature of the isointensity lines) are detected in all the images of the sequence. The trajectories are built by matching salient points of consecutive images whose neighbourhoods are coherent. Our first contribution consists in the analysis of the trajectories on a group of pictures, which improves the motion estimation quality. Moreover, we use a space-time weighting for each trajectory which makes it possible to add a temporal constraint on the movement while taking into account the local geometrical deformations of the object ignored by a global motion model. The second method performs a space-time segmentation. The object contour motion is estimated using the information contained in an outer-layer centered on the object contour. Our first contribution is the use of this outer-layer which contains information about both the background and the object in a local context. The matching using a statistical similarity measure (residual entropy) allows to improve the tracking while facilitating the choice of the optimal size of the crown. Finally, we propose a fast implementation of an existing tracking method of region of interest. This method relies on the use of a statistical similarity measure : the Kullback-Leibler divergence. This divergence can be estimated in a high dimension space using k-th nearest neighbors distance. These calculations being computationally very expensive, we propose a parallel implementation of the exhaustive search of the k-th nearest neighbors using GPU programming (via the programming interface NVIDIA CUDA). We show that this implementation speeds-up the tracking process up to a factor 15 compared to a classical implementation of this search using data structuring methods
YASSINE, ALI. „De la localisation et du suivi par vision monoculaire d'objets polyedriques articules modelises“. Clermont-Ferrand 2, 1995. http://www.theses.fr/1995CLF21753.
Der volle Inhalt der QuelleCarminati, Lionel. „Détection et suivi d'objets dans les scènes animées : application à la vidéo surveillance“. Bordeaux 1, 2006. http://www.theses.fr/2006BOR13190.
Der volle Inhalt der QuellePressigout, Muriel Marchand Éric. „Approches hybrides pour le suivi temps-réel d'objets complexes dans des séquences vidéos“. [S.l.] : [s.n.], 2006. ftp://ftp.irisa.fr/techreports/theses/2006/pressigout.pdf.
Der volle Inhalt der QuelleIzquierdo, David. „Contribution au développement d'une architecture générique dédiée au suivi d'objets en télésurveillance : application au suivi de véhicules et de visages“. Bordeaux 1, 2004. http://www.theses.fr/2004BOR12889.
Der volle Inhalt der QuelleDehais, Christophe. „Contributions pour les applications de réalité augmentée : suivi visuel et recalage 2D. Suivi d'objets 3D représentés par des modèles par points“. Phd thesis, Toulouse, INPT, 2008. http://oatao.univ-toulouse.fr/7244/1/dehais.pdf.
Der volle Inhalt der QuelleGarcia, Vincent. „Suivi d'objets d'intérêt dans une séquence d`images : des points saillants aux mesures statistiques“. Phd thesis, Université de Nice Sophia-Antipolis, 2008. http://tel.archives-ouvertes.fr/tel-00380372.
Der volle Inhalt der QuelleBonnaud, Laurent. „Schémas de suivi d'objets vidéo dans une séquence animée : application à l'interpolation d'images intermédiaires“. Phd thesis, Université Rennes 1, 1998. http://tel.archives-ouvertes.fr/tel-00070533.
Der volle Inhalt der Quelleaux séquences d'images, pour des applications multimédia. Ce travail est
divisé en deux contributions principales~: un algorithme de segmentation
d'images en objets vidéo en mouvement, et une méthode d'interpolation
temporelle opérant sur ces objets.
La segmentation de la séquence est effectuée par un algorithme de suivi
temporel. Un algorithme de segmentation spatio-temporelle est utilisé
initialement pour obtenir des régions dans la première image de la séquence.
Cette partition est ensuite suivie par une technique de contours actifs, qui
opère sur une nouvelle représentation de la segmentation, composée des
frontières ouvertes séparant les régions. L'algorithme estime à la fois le
mouvement des frontières et celui des régions. Il est capable de suivre
plusieurs objets simultanément et de traiter les occultations entre eux. Des
résultats, obtenus sur des séquences d'images réelles, montrent que cet
algorithme permet une bonne stabilité temporelle de la segmentation et une
bonne précision des frontières.
Le but de l'algorithme d'interpolation est de reconstruire des images
intermédiaires entre deux images de la séquence. Il s'agit d'un algorithme de
faible complexité qui peut être utilisé à la fin d'une chaîne codeur/décodeur.
L'interpolation est compensée en mouvement et utilise le mouvement des régions,
estimé pendant la phase de suivi. Il est aussi basé objets, dans le sens où il
utilise la segmentation pour prédire correctement les zones d'occultation. Cet
algorithme peut être utilisé pour trois applications différentes~: le codage
interpolatif (où des images de la séquence sont prédites par interpolation),
l'adaptation de la fréquence de la séquence à la fréquence d'affichage du
terminal de visualisation dans une transmission multipoints et la
reconstruction d'images manquantes (où l'on calcule des images non observées).
Des résultats expérimentaux pour la première application montrent que pour une
qualité de reconstruction donnée, la taux de compression moyen sur un groupe
d'images est plus élevé en utilisant l'interpolation qu'avec une prédiction
causale.
Ait, Fares Wassima. „Détection et suivi d'objets par vision fondés sur segmentation par contour actif basé région“. Phd thesis, Université Paul Sabatier - Toulouse III, 2013. http://tel.archives-ouvertes.fr/tel-00932263.
Der volle Inhalt der QuelleBonnaud, Laurent. „Schemas de suivi d'objets video dans une sequence animee : application a l'interpolation d'images intermediaires“. Rennes 1, 1998. http://www.theses.fr/1998REN10149.
Der volle Inhalt der QuelleNguyen, Xuan Son. „Exploitation de réseaux bayésiens dynamiques et du filtre particulaire pour le suivi d'objets articulés“. Paris 6, 2013. http://www.theses.fr/2013PA066145.
Der volle Inhalt der QuelleArticulated object tracking has now become a very active research area inthe field of computer vision. One of its applications, i. E. Human track-ing, is used in a variety of domains, such as security surveillance, humancomputer interface, gait analysis,. . . The problem is also of interest from thetheoretical point of view. Some of its challenges include, for example, thehigh dimensionality of state spaces, self-occlusions, kinematic ambiguities orsingularities, making it hard to solve and hence, attractive for the trackingcommunity. Particle Filter (PF) has been shown to be an effective method for solving visual tracking problems. This is due to its ability to deal with non-linear, non-Gaussian and multimodal distributions encountered in such problems. The key idea of particle filter is to approximate the posterior distribution of the target object state by a set of weighted samples. These samples evolve using a proposal distribution and their weights are updated by involving new observations. Under some assumptions, it can be shown that the distribution estimated by particle filter converges in a statistical sense to thetarget distribution. Unfortunately, in high dimensional problems, such asarticulated object tracking problems, the number of samples required for ap-proximating the target distribution can be prohibitively large since it growsexponentially with the number of dimensions (e. G. , the number of partsof the object), making the particle filter impractical. To reduce the com-plexity of tracking algorithms in such problems, various methods have beenproposed. One family of approaches that has attracted many researchers isbased on the decomposition of the state space into smaller dimensional sub-spaces where tracking can be achieved using classical methods. This resultsin tracking algorithms that are linear instead of exponential in the numberof parts of the object. Bayesian Networks (BN) offer a very effective way to represent articulated objects and to express the relationship between their different partssince the object can be naturally modeled by graphs where each part of theobject is represented by a node and the physical link between two neighborparts is represented by an edge. Most of conditional independence relation-ships induced by the structural constraints of the articulated objects can beeasily encoded in BN. This kind of graphical model has been exploited forarticulated object tracking in many works and has been shown to be a pow-erful tool for modeling the tracking problem in decomposition approaches. In this thesis, we focus on articulated object tracking and decompositiontechniques to deal with the high dimensionality of the state space describingthe problem. Our first approach is based on the state-of-the-art algorithmfor articulated object tracking: Partition Sampling (PS). First, we developan algorithm called Swapping-Based Partitioned Sampling (SBPS). In thisalgorithm, the prediction/correction step of PF is performed for a groupof parts in parallel instead of part after part as in PS. We also introducean operation called swapping which produces better particles, i. E. , particlesthat are nearer to the modes of the target distribution, after the correctionstep of the algorithm. We provide a principled way to select the set of partsprocessed in parallel and to perform the swapping operation so that the pos-terior distribution is correctly estimated. This approach enables to reducethe number of resampling steps of PS and to increase the tracking accu-racy due to the higher number of particles near the modes of the posteriordistribution obtained by the swapping operation. Because the swapping operation generates more particles near the modesof the posterior distribution, this might lead to a situation where the poste-rior distribution is represented by only a few distinct particles, that inducesa loss of particle diversity, resulting in tracking failure in some cases, e. G. When there is a sudden change in movements of the object. To address thisproblem, we introduce an algorithm called DBN-Based Combinatorial Re-sampling for articulated object tracking. Adding this resampling scheme intoa particle filter produces a new algorithm called Particle Filter with Combi-natorial Resampling (PFCR). Instead of aiming to find the best swapping,we create a particle set which contains particles generated from all possiblepermutations, implicitly constructed to avoid resampling over a particle setof exponential size. This approach allows increasing the number of parti-cles near the modes of the posterior distribution but also the diversity ofparticles as compared to SBPS, thus improving the tracking accuracy andreducing tracking failure. Our third approach for articulated object tracking, introduced in thisthesis, is based on a hierarchical search and Particle Swarm Optimization(PSO). This approach called Hierarchical Annealed Particle Swarm Opti-mization Particle Filter (HAPSOPF) aims to increase the tracking accuracyand reduce the computational cost of the tracking algorithm by integratingthe benefits of these two methods. First, the searching efficiency is improvedby performing PSO in subspaces whose dimension is much lower than thatof the original space and therefore the tracking accuracy is increased. Second, the search is performed in the same manner as PS, leading to a reducednumber of particles required for tracking. As a result, the computationalcost of the tracking algorithm is reduced. Moreover, some important factorsare introduced into the update equations of PSO to deal with the problemof noisy observation in articulated object tracking
Ait, Fares Wassima. „Détection et suivi d'objets par vision fondés sur segmentation par contour actif base région“. Toulouse 3, 2013. http://thesesups.ups-tlse.fr/2143/.
Der volle Inhalt der QuelleObject segmentation and tracking is a challenging area of ongoing research in computer vision. One important application lies in robotics where the ability to accurately segment an object of interest from its background is crucial and particularly on images acquired onboard during robot motion. Object segmentation technique consists in separating the object region from the image background according to a pre-defined criterion. Object tracking is a process of determining the positions of moving objects in image sequences. Several techniques can be applied to ensure these operations. In this thesis, we are interested to segment and track objects in video sequences using active contour method due to its robustness and efficiency to segment and track non-rigid objects. Active contour method consists in making a curve converge from an initial position around the object to be detected towards this object boundary according to a pre-defined criterion. We employ criteria which depend on the image regions what may impose certain constraints on the characteristics of these regions as a homogeneity assumption. This assumption may not always be verified due to the heterogeneity often present in images. In order to cope with the heterogeneity that may appear either in the object of interest or in the image background in noisy images using an inadequate active contour initialization, we propose a technique that combines local and global statistics in order to compute the segmentation criterion. By using a radius with a fixed size, a half-disk is superposed on each point of the active contour to define the local extraction regions. However, when the heterogeneity appears on both the object of interest and the image background, we develop a new technique based on a flexible radius that defines two half-disks with two different radius values to extract the local information. The choice of the value of these two radii is determined by taking into consideration the object size as well as the distance separating the object of interest from its neighbors. Finally, to track a mobile object within a video sequence using the active contour method, we develop a hybrid object tracking approach based on region characteristics and on motion vector of interest points extracted on the object region. Using our approach, the initial active contour for each image will be adequately adjusted in a way that it will be as close as possible to the actual boundary of the object of interest so that the evolution of active contour based on characteristics of the region will not be trapped by false contours. Simulation results on synthetic and real images validate the effectiveness of the proposed approaches
Minetto, Rodrigo. „Reconnaissance de zones de texte et suivi d'objets dans les images et les vidéos“. Paris 6, 2012. http://www.theses.fr/2012PA066108.
Der volle Inhalt der QuelleIn this thesis we address three computer vision problems: (1) the detection and recognition of flat text objects in images of real scenes; (2) the tracking of such text objects in a digital video; and (3) the tracking an arbitrary three-dimensional rigid object with known markings in a digital video. For each problem we developed innovative algorithms, which are at least as accurate and robust as other state-of-the-art algorithms. Specifically, for text recognition we developed (and extensively evaluated) a new HOG-based descriptor specialized for Roman script, which we call T-HOG, and showed its value as a post-filter for an existing text detector (SnooperText). We also improved the SnooperText algorithm by using the multi-scale technique to handle widely different letter sizes while limiting the sensitivity of the algorithm to various artifacts. For text tracking, we describe four basic ways of combining a text detector and a text tracker, and we developed a specific tracker based on a particle-filter which exploits the T-HOG recognizer. For rigid object tracking we developed a new accurate and robust algorithm (AFFTrack) that combines the KLT feature tracker with an improved camera calibration procedure. We extensively tested our algorithms on several benchmarks well-known in the literature. We also created benchmarks (publicly available) for the evaluation of text detection and tracking and rigid object tracking algorithms
Dewaele, Guillaume. „Modélisation, suivi et simulation d'objets articulés et déformables : application au modelage réel d'une argile virtuelle“. Phd thesis, Grenoble INPG, 2005. http://tel.archives-ouvertes.fr/tel-00584946.
Der volle Inhalt der QuelleDufour, Alexandre. „Segmentation, suivi et visualisation d'objets biologiques en microscopie 3D par fluorescence : Approches par modèles déformables“. Phd thesis, Université René Descartes - Paris V, 2007. http://tel.archives-ouvertes.fr/tel-00271191.
Der volle Inhalt der QuelleLes modèles déformables, également connus sous le nom de contours actifs, font actuellement partie des méthodes de pointe en analyse d'images pour la segmentation et le suivi d'objets grâce à leur robustesse, leur flexibilité et leur représentation à haut niveau sémantique des entités recherchées. Afin de les adapter à notre problématique, nous devons faire face à diverses difficultés. Tout d'abord, les méthodes existantes se réfèrent souvent aux variations locales d'intensité (ou gradients) de l'image pour détecter le contour des objets recherchés. Cette approche est inefficace en microscopie tridimensionnelle par fluorescence, où les gradients sont très peu prononcés selon l'axe de profondeur de l'image. Ensuite, nous devons gérer le suivi d'objets multiples susceptibles d'entrer en contact en évitant leur confusion. Enfin, nous devons mettre en place un système permettant de visualiser efficacement les contours durant leur déformation sans altérer les temps de calcul.
Dans la première partie de ce travail, nous pallions à ces problèmes en proposant un modèle de segmentation et de suivi multi-objets basé sur le formalisme des lignes de niveaux (ou level sets) et exploitant la fonctionnelle de Mumford et Shah. La méthode obtenue donne des résultats quantitatifs satisfaisants, mais ne se prête pas efficacement au rendu 3D de la scène, pour lequel nous sommes tributaires d'algorithmes dédiés à la reconstruction 3D (e.g. la méthode des "Marching Cubes"), souvent coûteux en mémoire et en temps de calcul. De plus, ces algorithmes peuvent induire des erreurs d'approximation et ainsi entraîner une mauvaise interprétation des résultats.
Dans la seconde partie, nous proposons une variation de la méthode précédente en remplaçant le formalisme des lignes de niveaux par celui des maillages triangulaires, très populaire dans le domaine de la conception assistée par ordinateur (CAO) pour leur rendu 3D rapide et précis. Cette nouvelle approche produit des résultats quantitatifs équivalents, en revanche le formalisme des maillages permet d'une part de réduire considérablement la complexité du problème et autorise d'autre part à effectuer un rendu 3D précis de la scène parallèlement au processus de segmentation, réduisant d'autant plus les temps de calculs.
Les performances des deux méthodes proposées sont d'abord évaluées puis comparées sur un jeu de données simulées reproduisant le mieux possible les caractéristiques des images réelles. Ensuite, nous nous intéressons plus particulièrement à l'évaluation de la méthode par maillages sur des données réelles, en évaluant la robustesse et la stabilité de quelques descripteurs de forme simples sur des expériences d'imagerie haut-débit. Enfin, nous présentons des applications concrètes de la méthode à des problématiques biologiques réelles, réalisées en collaboration avec d'autres équipes de l'Institut Pasteur de Corée.
Benamara, Mohamed Adel. „Suivi visuel d'objets dans un réseau de caméras intelligentes : application au systèmes de manutention automatisés“. Thesis, Lyon, 2018. http://www.theses.fr/2018LYSE2136.
Der volle Inhalt der QuelleIntralogistics (or internal logistics) focuses on the management and optimization of internal production and distribution processes within warehouses, distribution centers, and factories. Automated handling systems play a crucial role in the internal logistics of several industries such as e-commerce, postal messaging, retail, manufacturing, airport transport, etc. These systems are composed by multiple high-speed conveyor lines that provide safe and reliable transportation of a large volume of goods and merchandise while reducing costs.The automation of the conveying process relies on the identification and the real-time tracking of the transported loads. In this thesis, we designed a tracking solution that employs a network of smart cameras with an overlapping field of view. The goal is to provide tracking information to control an automated handling system.Multiple object tracking is a fundamental problem of computer vision that has many applications such as video surveillance, robotics, autonomous cars, etc. We integrated several building blocks traditionally applied to traffic surveillance or human activities monitoring to constitute a tracking pipeline. We used this baseline tracking pipeline to characterize contextual scene information proper to the conveying scenario. We integrated this contextual information to the tracking pipeline to enhance the performance. In particular, we took into account the state of moving objects that become stationary in the background subtraction step to prevent their absorption to the background model. We have also exploited the regularity of objects trajectory to enhance the motion model associated with the tracked objects. Finally, we integrated the precedence ordering constraint among the conveyed object to reidentify them when they are close to each other.We have also tackled practical problems related to the optimization the execution of the proposed tracking problem in the multi-core architectures of smart cameras. In particular, we proposed a dynamic learning process that extracts the region of the image that corresponds to the conveyor lines. We reduced the number of the processed pixel by restricting the processing to this region of interest. We also proposed a parallelization strategy that adaptively partitions this region of interest of the image, in order to balance the workload between the different cores of the smart cameras.Finally, we proposed a multiple cameras tracking algorithms based on event composition. This approach fuses the local tracking generated by the smart cameras to form global object trajectories and information from third party systems such as the destination of the object entered by operators on a terminal. We validated the proposed approach for the control of a sorting system deployed in a postal distribution warehouse. A network of cameras composed of 32 cameras tracks more than 400.000 parcel/day in injections lines. The tracking error rate is less than 1 parcel in a 1000 (0.1%)
Andries, Mihai. „Localisation et suivi d'humains et d'objets, et contrôle de robots au travers d'un sol sensible“. Thesis, Université de Lorraine, 2015. http://www.theses.fr/2015LORR0293.
Der volle Inhalt der QuelleThis thesis explores the capabilities of an ambient intelligence equipped with a load-sensing floor. It deals with the problem of perceiving the environment through a network of low-resolution sensors. Challenges include the interpretation of spread loads for objects with multiple points of support, weight ambiguities between objects, variation of persons’ weight during dynamic activities, etc. We introduce new techniques, partly inspired from the field of computer vision, for detecting, tracking and recognizing the entities located on the floor. We also introduce new modes of interaction between environments equipped with such floor sensors and robots evolving inside them. This enables non-intrusive interpretation of events happening inside environments with embedded ambient intelligence, with applications in assisted living, senile care, continuous health diagnosis, home security, and robotic navigation