Robotic technology is advancing out of the laboratory and into the everyday world. This world is less ordered than the laboratory and requires an increased ability to identify, target, and track objects of importance. The Bayes filter is the ideal algorithm for tracking a single target and there exists a significant body of work detailing tractable approximations of it with the notable examples of the Kalman and Extended Kalman filter. Multiple target tracking also relies on a similar principle and the Kalman and Extended Kalman filter have multi-target implementations as well. Other method include the PHD filter and Multiple Hypothesis tracker. One issue is that these methods were formulated to only track one classification of target. With the increased need for robust perception, there exists a need to develop a target tracking algorithm that is capable of identifying and tracking targets of multiple classifications. This thesis examines two of these methods: the Probability Hypothesis Density (PHD) filter and the Multiple Hypothesis Tracker (MHT). A Matlab-based simulation of an office floor plan is developed and a simulation UGV equipped with a camera is set the task of navigating the floor plan and identifying targets. Results of these experiments indicated that both methods are mathematically capable of achieving this. However, there was a significant reliance on post-processing to verify the performance of each algorithm and filter out noisy sensor inputs indicating that specific multi-target multi-class implementations of each algorithm should be implemented with a detailed and more accurate sensor model. / Mechanical Engineering

Les débris spatiaux sont devenus une menace considérable pour la viabilité des satellites opérationnels en orbite basse. Afin de pouvoir éviter des collisions accidentelles, des systèmes de surveillance de l'espace existent mais sont limités en performances de détection pour les objets de petite taille (diamètre inférieur à 10cm), ce qui pousse à l'étude de nouvelles solutions. Cette thèse a pour objectif d'appuyer la faisabilité d'un système radar au sol utilisant un champ de veille étroit pour le catalogage de petits débris en orbite basse. Un tel système fournirait en effet des observations dites « isolées », c'est-à-dire qu'une orbite n'est pas immédiatement déductible de chacune d'entre elles. Le grand nombre combinaisons nécessaires est alors prohibitif en termes de temps de calcul pour la résolution de ce problème de pistage. Nous proposons dans ces travaux une nouvelle méthode pour initialiser les pistes, c'est-à-dire associer des observations isolées avec une faible ambiguïté et en déduire des orbites précises. Les pistes ainsi obtenues sont combinées et filtrées grâce à un algorithme de pistage multicible que nous avons adapté aux particularités du problème. Avec un taux de couverture de plus de 80 % obtenu en temps réel sur 3 jours pour des scénarios de 500 à 800 objets en plus d'un fort taux de fausses alarmes, les performances de la méthode proposée tendent à prouver la faisabilité du système envisagé. Afin d'extrapoler les résultats obtenus à de plus fortes densités d'observations, nous proposons un modèle de complexité combinatoire calibré sur les performances de l'algorithme aux faibles densités. L'apport d'un second capteur identique est également étudié et met en évidence un point de compromis entre réactivité et complexité combinatoire, ce qui offre un degré de liberté supplémentaire dans la conception d'un tel système. / Space debris have become a significant threat to the viability of operational satellites in Low-Earth-Orbit. In order to avoid accidental collisions, space surveillance systems exist but their detection performance is limited for the small debris (less than 10cm). New solutions are then at study. This thesis aims at supporting the feasibility of a ground-based radar sensor with a narrow-fence type field of regard for the cataloging of the small space debris. Such a system would produce “isolated” observations, that is to say that an orbit is not directly available from each one of them. The large number of potential combinations is then computationally prohibitive for solving this tracking problem. In this work, we propose a new method for track initiation, i.e. associating isolated observations with little ambiguity and deduce accurate orbits. The obtained set of tracks are combined and filtered using an multitarget tracking algorithm that we have adapted to the peculiarities of the problem. With a coverage rate of more than 80% in real-time on 3 days for 500 to 800-objects scenarios in addition of a high false alarm rate, the performance of the proposed method supports the feasibility of the considered system. Aiming at extrapolating the obtained results to higher observation densities, we propose a combinatorial complexity model calibrated with the algorithm performance for low detection densities. The contribution of a second identical sensor is also assessed and reveals a possible trade-off between reactivity and combinatorial complexity, which offers an additional degree of freedom in the design of such a system.

This master's thesis prepared the programme which is able to follow the trajectories of the movement of people and based on this to create various statistics. In practice it is an effective marketing tool which can be used for instance for customer flow analyses, optimal evaluation of opening hours, visitor traffic analyses and for a lot of other benefits. Histograms of oriented gradients, SVM classificator and optical flow monitoring were used to solve this problem. The method of multiple hypothesis tracking was selected for the association data. The system's quality was evaluated from the video footage of the street with the large concentration of pedestrians and from the school's camera system, where the movement in the corridor was monitored and the number of people counted.

Nous nous intéressons dans cette thèse à la vision à travers les murs (VTM) par radar ULB, avec comme objectif la mise au point d’une chaîne de traitement de l’information (CTI) complète pouvant être utilisée par différents types de radar imageur VTM. Pour ce faire, nous souhaitons prendre en compte le moins possible d’information a priori, ni sur les cibles, ni sur leur contexte environnemental. De plus, la CTI doit répondre à des critères d’adaptabilité et de modularité pour pouvoir traiter les informations issues de deux types de radar, notamment, le pulsé et le FMCW, développés dans deux projets dans lesquels s’inscrivent les travaux de cette thèse. L’imagerie radar est un point important dans ce contexte, nous l’abordons par la combinaison des algorithmes de rétroprojection et trilatération, et montrons l’amélioration apportée avec l’utilisation d’un détecteur TFAC prenant en compte la forme des signatures des cibles. La mise au point de la CTI est notre principale contribution. Le flux d’images radar obtenu est scindé en deux parties. La première séquence dynamique contient les cibles mobiles qui sont ensuite suivies par une approche multihypothèse. La seconde séquence statique contient les cibles stationnaires ainsi que les murs intérieurs qui sont détectés par une méthode s’appuyant sur la transformée de Radon. Nous avons produit un simulateur VTM fonctionnant dans le domaine temporel et fréquentiel pour mettre au point les algorithmes de la CTI et tester leur robustesse. Plusieurs scénarios de simulation ainsi que de mesures expérimentales, montrent que la CTI construite est pertinente et robuste. Elle est ainsi validée pour les deux systèmes radar. / This 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.

Micro-unmanned aerial rotorcraft are quickly gaining acceptance as indoor platforms for performing stealth, surveillance, and rescue and reconnaissance missions. These rotorcraft are generally required to operate in cluttered, unknown, and dynamic GPS-denied environments, which present threats to the safe operation of the vehicle. To overcome these environmental challenges, we describe a system that is capable of localizing itself by producing accurate odometry estimates that can detect and track moving objects and avoid collisions with obstacles while following a moving target using a laser range finder. Our system has been implemented in the Simulink environment in MATLAB. Various simulations have shown our methods to work well, even in the presence of sensor noise and out-of-plane motion. Our system is capable of localizing itself within ±20 mm in North and East and ±0.5 degrees in ψ while detecting and tracking

Le suivi de particules est une méthode de choix pour comprendre les mécanismes intra-cellulaires car il fournit des moyens robustes et précis de caractériser la dynamiques des objets mobiles à l'échelle micro et nano métrique. Cette thèse traite de plusieurs aspects liés au problème du suivi de plusieurs centaines de particules dans des conditions bruitées. Nous présentons des techniques nouvelles basée sur des méthodes mathématiques robustes qui nous permettent des suivre des particules sous-résolutives dans les conditions variées qui sont rencontrées en imagerie cellulaire. Détection de particules : nous avons tout d'abord traité le problème de la détection de particules dans les images fluorescentes contenant un fond structuré. L'idée clé de la méthode est l'utilisation d'une technique de séparation de sources : l'algorithme d'Analyse en Composantes Morphologiques (ACM), pour séparer le fond des particules en exploitant leur différence de morphologie dans les images. Nous avons effectué un certain nombre de modifications à l'ACM pour l'adapter aux caractéristiques des images biologiques en fluorescence. Par exemple, nous avons proposé l'utilisation du dictionnaire de Curvelet et d'un dictionnaire de d'ondelettes, avec des à priori de parcimonie différents, afin de séparer le signal des particules du fond. Une fois la séparation de sources effectuée, l'image sans fond peut être analysée pour identifier de manière robuste la position des particules et pour les suivre au cours du temps. Modélisation du problème de suivi : nous avons proposé un cadre de travail statistique global qui tient compte des nombreux aspects du problème de suivi de particules dans des conditions bruitées. Le cadre de travail probabiliste que nous avons mis au point contient de nombreux modèles qui sont dédiés à l'imagerie biologique, tels que des modèles statistiques de mouvement des particules en milieu cellulaire. Nous avons aussi défini la concept de perceiability d'une cible dans le cas des particules biologiques. Grâce à ce modèle l'existence d'une particule est explicitement modélisée et quantifiée, ce qui nous permet de résoudre les problèmes de création et de terminaison des trajectoires au sein même de notre cadre probabiliste de suivi. Le cadre de travail proposé bénéficie d'une grande flexibilité mais reste facile à adapter car chaque paramètre du modèle trouve une interprétation simple et intuitive. Ainsi, notre modèle probabiliste de suivi nous a permis de modéliser de manière exhaustive un grand nombre de systèmes biologiques différents. Mise au point d'un algorithme de suivi : nous avons reformulé l'algorithme de suivi nommé Multiple Hypothesis Tracking (MHT) pour qu'il inclue notre modèle probabiliste de suivi dédié aux particules biologiques, et nous avons proposé une implémentation rapide qui permet de suivre de nombreuses particules dans des conditions d'imagerie dégradées. L'\textit{Enhanced} MHT (E-MHT) que nous avons proposé tire pleinement partie du modèle de suivi en incorporant la connaissance des images futures, ce qui augmente significativement le pouvoir discriminant des critères statistiques. En conséquence, l'E-MHT est capable d'identifier automatiquement les détections erronée et de détecter les événements d'apparition et de disparition des particules. Nous avons résolu le problème de la complexité de la tache de suivi grâce à un design de l'algorithme que exploite la topologie en arbre des solution et à la possibilité d'effectuer les calculs de manière parallèle. Une série de tests comparatifs entre l'E-MHT et des méthodes existantes de suivi a été réalisée avec des séquences d'images synthétiques 2D et avec des jeux de données réels 2D et 3D. Dans chaque cas l'E-MHT a montré des performances supérieures par rapport aux méthodes standards, avec une capacité remarquable à supporter des conditions d'imagerie très dégradées. Nous avons appliqué les méthodes de suivi proposées dans le cadre de plusieurs projets biologiques, ce qui a conduit à des résultats biologiques originaux. La flexibilité et la robustesse de notre méthode nous a notamment permis de suivre des prions infectant des cellules, de caractériser le transport de protéines lors du développement de l'ovocyte de la drosophile, ainsi que d'étudier la trafic d'ARN messager dans l'ovocyte de drosophile.

A cruise missile is a guided missile that uses a lifting wing and a jet propulsion system to allow sustained flight. Cruise missiles are, in essence, unmanned aircraft and they are generally designed to carry a large conventional or nuclear warhead many hundreds of miles with excellent accuracy. In this study, navigation and control studies on cruise missiles are performed. Due to the variety and complexity of the subsystems of the cruise missiles, the main concern is limited with the navigation system. Navigation system determines the position, velocity, attitude and time solutions of the missile. Therefore, it can be concluded that an accurate self-contained navigation system directly influences the success of the missile. In the study, modern radar data association algorithms are implemented as new Terrain Aided Navigation (TAN) algorithms which can be used with low-cost Inertial Measurement Units (IMU&rsquo / s). In order to perform the study, first a thorough survey of the literature on mid-course navigation of cruise missiles is performed. Then, study on modern radar data association algorithms and their implementations to TAN are done with simple simulations. At the case study part, a six degree of freedom (6 DOF) flight simulation tool is developed which includes the aerodynamic and dynamic model of the cruise missile model including error model of the navigation system. Finally, the performances of the designed navigation systems with the implemented TAN algorithms are examined in detail with the help of the simulations performed.

Multiple target tracking (MTT) is the process of identifying the number of targets present in a surveillance region and the state estimates, or track, of each target. MTT remains a challenging problem due to the NP-hard data association step, where unlabeled measurements are identified as either a measurement of an existing target, a new target, or a spurious measurement called clutter. Existing techniques suffer from at least one of the following drawbacks: divergence in clutter, underlying assumptions on the number of targets, high computational complexity, time-consuming implementation, poor performance at low detection rates, and/or poor track continuity. Our goal is to develop an efficient MTT algorithm that is simple yet effective and that maintains track continuity enabling persistent tracking of an unknown number of targets. A related field to tracking is regression analysis, where the parameters of static signals are estimated from a batch or a sequence of data. The random sample consensus (RANSAC) algorithm was developed to mitigate the effects of spurious measurements, and has since found wide application within the computer vision community due to its robustness and efficiency. The main concept of RANSAC is to form numerous simple hypotheses from a batch of data and identify the hypothesis with the most supporting measurements. Unfortunately, RANSAC is not designed to track multiple targets using sequential measurements.To this end, we have developed the recursive-RANSAC (R-RANSAC) algorithm, which tracks multiple signals in clutter without requiring prior knowledge of the number of existing signals. The basic premise of the R-RANSAC algorithm is to store a set of RANSAC hypotheses between time steps. New measurements are used to either update existing hypotheses or generate new hypotheses using RANSAC. Storing multiple hypotheses enables R-RANSAC to track multiple targets. Good tracks are identified when a sufficient number of measurements support a hypothesis track. The complexity of R-RANSAC is shown to be squared in the number of measurements and stored tracks, and under moderate assumptions R-RANSAC converges in mean to the true states. We apply R-RANSAC to a variety of simulation, camera, and radar tracking examples.

Indoor localization remains an open problem in geolocation research, and once this is solved the localization enables counting and tracking of building occupants. This information is vital in an emergency, enables occupancy-optimized heating or cooling, and assists smart buildings in tailoring services for occupants. Unfortunately, two prevalent technologies---GPS and cellular-based positioning---perform poorly indoors due to attenuation and multipath from the building. To address this issue, the research community devised many alternatives for indoor localization (e.g., beacons, RFID tags, Wi-Fi fingerprinting, and UWB to cite just a few examples). A drawback with most is the requirement for those being located to carry a properly-configured device at all times. An alternative based on computer vision techniques poses significant privacy concerns due to cameras recording building occupants. By contrast, ILoViT research makes novel use of accelerometers already present in some buildings. These sensors were originally intended to monitor structural health or to study structural dynamics. The key idea is that when a person's footstep-generated floor vibrations can be detected and located then it becomes possible to locate persons moving within a building. Vibration propagation in buildings has complexities not encountered by acoustic or radio wave propagation in air; thus, conventional localization algorithms are inadequate. ILoVIT algorithms account for these conditions and have been demonstrated in a public building to provide sub-meter accuracy. Localization provides the foundation for counting and tracking, but providing these additional capabilities confronts new challenges. In particular, how does one determine the correct association of footsteps to the person making them? The ILoViT research created two methods for solving the data association problem. One method only provides occupancy counting but has modest, polynomial time complexity. The other method draws inspiration from prior work in the radar community on the multi-target tracking problem, specifically drawing from the multiple hypothesis tracking strategy. This dissertation research makes new enhancements to this tracking strategy to account for human gait and characteristics of footstep-derived multilateration. The Virginia Polytechnic Institute and State University's College of Engineering recognized this dissertation research with the Paul E. Torgersen Graduate Student Research Excellence Award. / Ph. D. / Indoor localization remains an open problem in geolocation research, and once this is solved the localization enables counting and tracking of building occupants. This information is vital in an emergency, enables occupancy-optimized heating or cooling and assists smart buildings in tailoring services for occupants. Unfortunately, two prevalent technologies—GPS and cellular-based positioning—are ill-suited here due to the way a building’s weakens and distorts wireless signals. To address this issue the research community devised many alternatives for indoor localization. A drawback with most is the requirement for those being located to carry a properly-configured device at all times. An alternative based on computer vision techniques poses significant privacy concerns due to cameras recording building occupants. By contrast, ILoViT research makes novel use of a mature sensor technology already present in some buildings. These sensors were originally intended to monitor structural health or to study structural dynamics. The key idea behind this unconventional role for building sensors is that when a person’s footstep-generated floor vibrations can be detected and located then it is possible to locate persons moving within a building. Vibration propagation in buildings has complexities not encountered by acoustic or radio wave propagation in air; thus, conventional localization algorithms designed for those applications are inadequate. ILoVIT algorithms account for these conditions and have been demonstrated in a public building to provide sub-meter accuracy. Localization provides the foundation for counting and tracking, but providing these additional capabilities confronts new challenges. In particular, how does one determine the correct association of footsteps to the person making them? The ILoViT research created two methods for solving the data association problem. One method only provides area occupancy counting but has modest complexity. The other method draws inspiration from prior work in the radar community on the multi-target tracking problem, and the dissertation research makes new enhancements to account for human gait and footstep-based localization. The Virginia Polytechnic Institute and State University’s College of Engineering recognized this dissertation research with the Paul E. Torgersen Graduate Student Research Excellence Award.

<p>Signal processing within the neurophysiological field is challenging and requires short processing time and reliable results. In this thesis, three main problems are considered.</p><p>First, a modified line source model for simulation of muscle action potentials (APs) is presented. It is formulated in continuous-time as a convolution of a muscle-fiber dependent transmembrane current and an electrode dependent weighting (impedance) function. In the discretization of the model, the Nyquist criterion is addressed. By applying anti-aliasing filtering, it is possible to decrease the discretization frequency while retaining the accuracy. Finite length muscle fibers are incorporated in the model through a simple transformation of the weighting function. The presented model is suitable for modeling large motor units.</p><p>Second, the possibility of discerning the individual AP components of the concentric needle electromyogram (EMG) is explored. Simulated motor unit APs (MUAPs) are prefiltered using Wiener filtering. The mean fiber concentration (MFC) and jitter are estimated from the prefiltered MUAPs. The results indicate that the assessment of the MFC may well benefit from the presented approach and that the jitter may be estimated from the concentric needle EMG with an accuracy comparable with traditional single fiber EMG.</p><p>Third, automatic, rather than manual, detection and discrimination of recorded C-fiber APs is addressed. The algorithm, detects the Aps reliably using a matched filter. Then, the detected APs are discriminated using multiple hypothesis tracking combined with Kalman filtering which identifies the APs originating from the same C-fiber. To improve the performance, an amplitude estimate is incorporated into the tracking algorithm. Several years of use show that the performance of the algorithm is excellent with minimal need for audit.</p>