Kliknij ten link, aby zobaczyć inne rodzaje publikacji na ten temat: Pedestrian crowds.
Rozprawy doktorskie na temat „Pedestrian crowds”
Utwórz poprawne odniesienie w stylach APA, MLA, Chicago, Harvard i wielu innych
Sprawdź 44 najlepszych rozpraw doktorskich naukowych na temat „Pedestrian crowds”.
Przycisk „Dodaj do bibliografii” jest dostępny obok każdej pracy w bibliografii. Użyj go – a my automatycznie utworzymy odniesienie bibliograficzne do wybranej pracy w stylu cytowania, którego potrzebujesz: APA, MLA, Harvard, Chicago, Vancouver itp.
Możesz również pobrać pełny tekst publikacji naukowej w formacie „.pdf” i przeczytać adnotację do pracy online, jeśli odpowiednie parametry są dostępne w metadanych.
Przeglądaj rozprawy doktorskie z różnych dziedzin i twórz odpowiednie bibliografie.
1
Johansson, Anders. "Data-Driven Modeling of Pedestrian Crowds". Doctoral thesis, Saechsische Landesbibliothek- Staats- und Universitaetsbibliothek Dresden, 2009. http://nbn-resolving.de/urn:nbn:de:bsz:14-qucosa-20900.
Pełny tekst źródłaStreszczenie:
At the starting point of the work leading to this doctoral thesis, in January 2005, the work on pedestrians was almost exclusively oriented towards computer simulations and on evacuation experiments. Since then, there have been many studies on new methods for extracting empirical data of pedestrian movements (mainly based on video analysis, lasers, and infrared cameras), but most of the work is still focused on artificial setups for crowds moving through corridors and crowds passing bottlenecks. Even though these controlled experiments are important to understand crowd dynamics, there is a knowledge gap between these experiments and the understanding of the dynamics leading to and occurring during large crowd disasters, when sometimes hundreds of thousands or even millions of pedestrians are involved. Numerous crowd disasters occur every year at large gatherings around the world. Unfortunately, the information about the (spatio-temporal) development of these events tend to be qualitative rather than quantitative. Video recordings from the crowd disaster in Mina, Kingdom of Saudi Arabia, on the 12th of January 2006, where hundreds of pilgrims lost their lives during the annual Muslim pilgrimage to Makkah, gave the possibility to scientifically evaluate the dynamics of the crowd. With this video material, it was possible to observe and analyze the behavior of the crowd under increasing crowd density, leading to the disaster. Based on the insights from the analysis of the crowd disaster described above, new tools and measures to detect and avoid critical crowd conditions have been proposed, and some of them have been implemented in order to reduce the likelihood of similar disasters in the future. Further contributions of this thesis are to empirically evaluate many of the previous assumptions used for pedestrian simulations. These assumptions are: * A pedestrian avoids collisions by changing her or his walking speed by an acceleration which is exponentially decaying with the distance to the pedestrian or object being avoided. * A pedestrian reacts stronger to what happens in front of her or him, than to what happens behind the back. * The movement of a crowd of pedestrians always follows a smooth flow-density relationship, called the fundamental diagram. * The walking speed will settle at 0 m/s at a specific maximum crowd density. The first two assumptions were found to be consistent with the data, but the pedestrian-flow theory had to be revised, since the two latter assumptions do not always hold. When these fundamental parts of pedestrian motion and avoiding maneuvers had been investigated, an improved version of the social-force-model was formulated. In order to enable the revision of previous works and the analysis of the crowd disaster mentioned above, algorithms used for video-tracking have been introduced. The novelty of this work concerns the uniqueness and quantity of data on which the algorithms are validated and calibrated, but also the focus on analyzing millions of pedestrians rather than hundreds. The aim of this thesis is to move from theoretical models and controlled lab conditions to applicable models for real-world conditionsAls diese Dissertation im Januar 2005 begonnen wurde, nutzten wissenschaftliche Untersuchungen von Fußgängern fast ausschließlich Computersimulationen und Evakuierungsexperimente. Seit dem haben viele Wissenschaftler an einer Verbesserung der Methoden gearbeitet. Heute werden empirische Daten mit Hilfe von Videoanalysen, Laser- und Infrarotkameras erhoben.Jedoch konzentrieren sich viele dieser Arbeiten auf künstliche Setups, in denen sich Fußgängermassen durch Korridore und Engpässe bewegen. Diese Experimente erlauben es, Massenbewegungen zu verstehen. Jedoch gibt es immer noch Forschungslücken. Es ist schwierig, unter solch kontrollierten Bedingungen Fortschritte darin zu erzielen, die auftretenden Dyamiken bei großen Katastrophen zu verstehen, in denen manchmal Hunderttausende oder sogar Millionen von Fußgängern involviert sind. Immer wieder kommt es zu Katastrophen in großen Menschenmengen. Leider sind von diesen Ereignissen häufig nur qualitative Informationen anstelle von quantitativen Daten erhältlich. Es ergab sich die besondere Gelegenheit, quantitatives Filmmaterial über eine Katastrophe in Mina (Königreich Saudi--Arabien) zu erhalten. Dort starben am 12. Januar 2006 hunderte von Pilgern während der jährlichen muslimischen Pilgerfahrt nach Mekka. Mit dem erhobenen Videomaterial konnte nachvollzogen werden, wie die Menschenmenge zuerst unbehindert fließen konnte, dann immer dichter wurde und wie es schließlich zur Katastrophe kam. Von den Erkenntnissen der Analyse der oben beschriebenen Katastrophe konnten neue Methoden entwickelt werden, die dabei helfen können, ähnliche Katastrophen in Zukunft zu vermeiden. Ein weiterer Beitrag dieser Dissertation besteht darin, einige Annahmen, die üblicherweise bei der Simulation von Fußgängerdynamiken gemacht werden, in Frage zu stellen und zu überarbeiten. Diese Annahmen sind: (1) Ein Fußgänger verhindert Zusammenstöße, indem er seine Schrittgeschwindigkeit so verändert, dass seine Beschleunigung exponentiell mit der Distanz zu dem zu umgehenden Fußgänger oder Objekt abnimmt. (2) Ein Fußgänger zeigt stärkere Reaktionen auf Ereignisse, die vor ihm passieren, als auf Ereignisse, die hinter ihm passieren. (3) Die Bewegung eines in einer Menschenmenge befindlichen Fußgängers folgt immer dem Strömungs--Dichte Verhältnis, was als Fundamental-Diagramm bezeichnet wird. (4) Die Laufgeschwindigkeit eines Fußgängers erreicht bei maximaler Menschendichte einem Wert von 0 m/s. Die ersten beiden Annahmen wurden von den empirischen Daten bestätigt. Unsere Analysen zeigen jedoch, dass die Annahmen 3 und 4 nicht immer gültig sind. Somit müssen Standardtheorien von Fußgängerdynamiken überarbeitet werden. Im Anschluß an die Analyse dieser fundamentalen Aspekte von Fußgängerverhalten und dem Verhalten bei Ausweichmanövern wird das Social-Force-Modell weiterentwickelt. Um auf vorhergehenden Arbeiten aufzubauen und um die oben beschriebene Katastrophe analysieren zu können, werden Algorithmen für die Video-Verfolgung von Fußgängerbewegungen entwickelt. Das Neue bei diesem Teil der Arbeit liegt nicht nur in dem verwendeten Verfahren selbst, sondern auch in der Einzigartigkeit und der großen Menge an verwendeten Daten, die mit diesem Verfahren analysiert werden. Ein zentrales Ziel dieser Arbeit besteht demnach in einer wissenschaftlichen Weiterentwicklung von theoretischen Modellen und kontrollierten Laborexperimenten hin zu Modellen, die unter realen Bedingungen tatsächlich anwendbar sind. Die Analyse von Fußgängern ist ein interdisziplinäres Feld, das von verschiedenen wissenschaftlichen Disziplinen mit verschiedenen Zielen betrieben wird. Leider gab es bislang wenig Bemühungen, die Resultate innerhalb dieser Teilgebiete im Rahmen einer konsistenten Theorie zu vereinen. Als seltene Ausnahmen können die Arbeiten von Teknomo und Antonini genannt werden. Diese Dissertation verfolgt das Ziel, diese theoretische Vereinigung weiter voran zu treiben. Dazu muss man zwischen der Neuerfindung des Rades und der Wiederverwendung nicht geprüfter Resultate abwägen. Dementsprechend ist ein Teil dieser Dissertation dem Vorhaben gewidmet, bisherige Forschung im Lichte empirischer Daten und neuer Methoden zu evaluieren. Da sich die Arbeit mit recht unterschiedlichen Aspekten von Fußgängerverhalten beschäftigt, konzentriert sich die Analyse in verschiedenen Teilen der Arbeit auf einige ausgewählte, alternative Modelle. Insbesondere bei der Modellierung und Simulation wird anstelle einer eingehenden Übersicht verschiedener Modelle eine Diskussion des speziellen Social-Force Modells präsentiert
2
Vandoni, Jennifer. "Ensemble Methods for Pedestrian Detection in Dense Crowds". Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLS116/document.
Pełny tekst źródłaStreszczenie:
Cette thèse s’intéresse à la détection des piétons dans des foules très denses depuis un système mono-camera, avec comme but d’obtenir des détections localisées de toutes les personnes. Ces détections peuvent être utilisées soit pour obtenir une estimation robuste de la densité, soit pour initialiser un algorithme de suivi. Les méthodologies classiques utilisées pour la détection de piétons s’adaptent mal au cas où seulement les têtes sont visibles, de part l’absence d’arrière-plan, l’homogénéité visuelle de la foule, la petite taille des objets et la présence d’occultations très fortes. En présence de problèmes difficiles tels que notre application, les approches à base d’apprentissage supervisé sont bien adaptées. Nous considérons un système à plusieurs classifieurs (Multiple Classifier System, MCS), composé de deux ensembles différents, le premier basé sur les classifieurs SVM (SVM- ensemble) et le deuxième basé sur les CNN (CNN-ensemble), combinés dans le cadre de la Théorie des Fonctions de Croyance (TFC). L’ensemble SVM est composé de plusieurs SVM exploitant les données issues d’un descripteur différent. La TFC nous permet de prendre en compte une valeur d’imprécision supposée correspondre soit à une imprécision dans la procédure de calibration, soit à une imprécision spatiale. Cependant, le manque de données labellisées pour le cas des foules très denses nuit à la génération d’ensembles de données d’entrainement et de validation robustes. Nous avons proposé un algorithme d’apprentissage actif de type Query-by- Committee (QBC) qui permet de sélectionner automatiquement de nouveaux échantillons d’apprentissage. Cet algorithme s’appuie sur des mesures évidentielles déduites des fonctions de croyance. Pour le second ensemble, pour exploiter les avancées de l’apprentissage profond, nous avons reformulé notre problème comme une tâche de segmentation en soft labels. Une architecture entièrement convolutionelle a été conçue pour détecter les petits objets grâce à des convolutions dilatées. Nous nous sommes appuyés sur la technique du dropout pour obtenir un ensemble CNN capable d’évaluer la fiabilité sur les prédictions du réseau lors de l’inférence. Les réalisations de cet ensemble sont ensuite combinées dans le cadre de la TFC. Pour conclure, nous montrons que la sortie du MCS peut être utile aussi pour le comptage de personnes. Nous avons proposé une méthodologie d’évaluation multi-échelle, très utile pour la communauté de modélisation car elle lie incertitude (probabilité d’erreur) et imprécision sur les valeurs de densité estiméesThis study deals with pedestrian detection in high- density crowds from a mono-camera system. The detections can be then used both to obtain robust density estimation, and to initialize a tracking algorithm. One of the most difficult challenges is that usual pedestrian detection methodologies do not scale well to high-density crowds, for reasons such as absence of background, high visual homogeneity, small size of the objects, and heavy occlusions. We cast the detection problem as a Multiple Classifier System (MCS), composed by two different ensembles of classifiers, the first one based on SVM (SVM-ensemble) and the second one based on CNN (CNN-ensemble), combined relying on the Belief Function Theory (BFT) to exploit their strengths for pixel-wise classification. SVM-ensemble is composed by several SVM detectors based on different gradient, texture and orientation descriptors, able to tackle the problem from different perspectives. BFT allows us to take into account the imprecision in addition to the uncertainty value provided by each classifier, which we consider coming from possible errors in the calibration procedure and from pixel neighbor's heterogeneity in the image space. However, scarcity of labeled data for specific dense crowd contexts reflects in the impossibility to obtain robust training and validation sets. By exploiting belief functions directly derived from the classifiers' combination, we propose an evidential Query-by-Committee (QBC) active learning algorithm to automatically select the most informative training samples. On the other side, we explore deep learning techniques by casting the problem as a segmentation task with soft labels, with a fully convolutional network designed to recover small objects thanks to a tailored use of dilated convolutions. In order to obtain a pixel-wise measure of reliability about the network's predictions, we create a CNN- ensemble by means of dropout at inference time, and we combine the different obtained realizations in the context of BFT. Finally, we show that the output map given by the MCS can be employed to perform people counting. We propose an evaluation method that can be applied at every scale, providing also uncertainty bounds on the estimated density
3
Berton, Florian. "Immersive virtual crowds : evaluation of pedestrian behaviours in virtual reality". Thesis, Rennes 1, 2020. http://www.theses.fr/2020REN1S056.
Pełny tekst źródłaStreszczenie:
La réalité virtuelle (RV) est devenu un outil de plus en plus utilisé afin d'étudier le comportement humain. En effet, son utilisation permet d'avoir un contrôle absolu sur les conditions expérimentales et de reproduire le même stimulus pour tous les participants. Dans cette thèse, nous utilisons la RV pour étudier le comportement piétons dans les foules afin par la suite d'améliorer les simulateurs de foules. En particulier nous nous intéressons à l'analyse couplée de la marche et du regard pour pouvoir comprendre et modéliser le voisinage d'interaction lors de la navigation. Dans nos premiers travaux, nous nous avons évalué l'impact de la RV sur l'activité du regard lors d’une interaction entre deux piétons, dans une étude où les participants réalisaient une tâche d'évitement de collision dans un environnement réel et virtuel. Par la suite nous nous sommes intéressés à une situation plus complexe qui est la navigation dans une rue peuplée. Nous avons de nouveau évalué l'impact de la RV sur l'activité du regard, puis nous nous sommes intéressé à l'impact de la densité de la foule sur cette activité. Finalement, dans une troisième étude nous avons simulé, en utilisant un rendu haptique, les collisions se produisant lors de la navigation dans une foule dense, et nous avons évalué l'influence de tel rendu sur la navigation des participants. En conclusion, nos résultats montrent que la réalité virtuelle est un outil pertinent pour l'étude du comportement des piétons dans les foules. En particulier, avec les récentes innovations technologiques, cet outil est adapté à l'étude de l'activité du regard, qui a d’ailleurs été peu explorée jusqu'à présent pour ce type de situationVirtual Reality (VR) has become more and more used as a tool to study human behaviour. Indeed, its use provides absolute control over experimental conditions and can reproduce the same stimulus for all participants. In this thesis, we use VR to investigate pedestrian behaviour in crowds in order to subsequently improve crowd simulators. In particular we are interested in a coupled analysis of locomotion and gaze in order to understand and model the interaction neighbourhood during navigation. In our first work, we evaluated the impact of VR on gaze activity during an interaction between two pedestrians, in a study where participants performed a collision avoidance task in a real and virtual environment. We then studied a more complex situation which is the navigation in a crowded street. We again evaluated the impact of VR on gaze activity and then explored the impact of crowd density on this activity. Finally, in a third study we simulated the collisions that occur when navigating in a dense crowd using haptic rendering, and evaluated the influence of such rendering on participants' locomotion. In conclusion, our results show that VR is a relevant tool to study pedestrian behaviour in crowds. In particular, with recent technological innovations, this tool is appropriate for the study of gaze activity, which to date has been little explored for this kind of situation
4
Feng, Weinan. "Multiple Human Body Detection in Crowds". Thesis, Högskolan i Gävle, Avdelningen för Industriell utveckling, IT och Samhällsbyggnad, 2012. http://urn.kb.se/resolve?urn=urn:nbn:se:hig:diva-12352.
Pełny tekst źródłaStreszczenie:
The objective of this project is to use digital imaging devices to monitor a delineated area of the public space and to register statistics about people moving across this area. A feasible detecting approach, which is based on background subtraction, has been developed and has been tested on 39 images. Individual pedestrians in images can be detected and counted. The approach is suitably used to detect and count pedestrians without overlapping. Accuracy rate of detection is higher than 80%.
5
Makmul, Juntima [Verfasser], i Simone [Akademischer Betreuer] Göttlich. "Microscopic and macroscopic models for pedestrian crowds / Juntima Makmul. Betreuer: Simone Göttlich". Mannheim : Universitätsbibliothek Mannheim, 2016. http://d-nb.info/1099910633/34.
Pełny tekst źródła
6
Bain, Nicolas. "Hydrodynamics of polarized crowds : experiments and theory". Thesis, Lyon, 2018. http://www.theses.fr/2018LYSEN078/document.
Pełny tekst źródłaStreszczenie:
Modéliser le mouvement des foules humaines est essentiel pour des situations aussi diverses que la prévention de risque dans les lieux publics, la planification d’évènements ou la création d’animations visuelles réalistes. Cependant, la difficulté de mener des expériences quantitatives limite notre compréhension de la dynamique des piétons, et le manque de mesures de référence rend impossible une comparaison poussée des modèles existants. Cette thèse tente d’augmenter notre compréhension des foules humaines par deux approches distinctes. Dans un premier temps, nous avons conduit une étude numérique et théorique pour étudier formation de lignes au sein de flux bidirectionnels d'agents motiles. Nous avons montré qu’une transition de phase critique du second ordre séparait un état mélangé d’un état constitué de lignes géantes le long desquelles se déplacent les agents visants une même direction. Cette séparation est caractéristique des systèmes actifs. Une approche hydrodynamique nous a ensuite permis de prouver que les phases mélangées sont aussi algébriquement corrélées dans la direction longitudinale. Nous avons expliqué et montré que ces fortes corrélations sont génériques de tous systèmes de flux bidirectionnels, qu’ils soient constitués de particules forcées ou de particules actives. Dans un second temps, nous avons mené une campagne expérimentale de grande envergure afin d’établir une expérience de référence des foules humaines. Nous avons pour cela choisi un système modèle, la zone d’attente de marathons. Dans ces foules de dizaines de milliers d’individus, nous avons quantitativement établi que les fluctuations de vitesse se propagent sur de grandes échelles, alors que les variations d’orientation s’évanouissent en quelques secondes. Grâce à ces mesures, nous avons construit une théorie prédictive hydrodynamique des foules polariséesModelling crowd motion is central to situations as diverse as risk prevention in mass events and visual effects rendering in the motion picture industry. The difficulty to perform quantitative measurements in model experiments, and the lack of reference experimental system, have however strongly limited our ability to model and control pedestrian flows. The aim of this thesis is to strengthen our understanding of human crowds, following two distinct approaches.First, we designed a numerical model to study the lane formation process among bidirectional flows of motile particles. We first evidenced the existence of two distinct phases: one fully laned and one homogeneously mixed, separated by a critical phase transition, unique to active systems. We then showed with a hydrodynamic approach that the mixed phase is algebraically correlated in the direction of the flow. We elucidated the origin of these strong correlations and proved that they were a universal feature of any system of oppositely moving particles, active of passive.Second, we conducted a substantial experimental campaign to establish a model experiment of human crowds. For that purpose we performed systematic measurements on crowds composed of tens of thousands of road-race participants in start corrals, a geometrically simple setup. We established that speed information propagates through polarized crowds over system spanning scales, while orientational information is lost in a few seconds. Building on these observations, we laid out a hydrodynamic theory of polarized crowds and demonstrated its predictive power
7
KHAN, SULTAN DAUD. "Automatic Detection and Computer Vision Analysis of Flow Dynamics and Social Groups in Pedestrian Crowds". Doctoral thesis, Università degli Studi di Milano-Bicocca, 2016. http://hdl.handle.net/10281/102644.
Pełny tekst źródłaStreszczenie:
Computer vision played a vital role in the field of video surveillance. However, recent developed computer vision algorithms rarely solve the problems related to real time crowd management. The phenomena of crowd like sports, festivals, concerts, political gatherings etc, are mostly observed in urban areas, which attracts hundreds of thousands people. In this thesis, we have developed algorithms that overcome some of the challenges encountered in videos of crowded environments such as sporting events, religious festivals, parades, concerts, train stations, airports, and malls. The main theme of this thesis is two fold ,i.e, understanding crowd dynamics in videos of (i), high density crowds and (ii) low density crowds. Typical examples of high density crowds include marathons, religious festivals while malls, airports, subways etc covers low dense situations. In this thesis, we adopt different approaches in order to deal with different kinds of problems coming from these two categories of crowd.
In particular, first part of the thesis, we adopt holistic approach to generate a global representation of the scene that captures both dynamics of the crowd and structure of the scene. This was achieved by extracting global features, i.e optical flow from the scene. For the crowd flow segmentation problem, the optical flows vectors are clustered by using K-means clustering followed by the blob absorption approach. Using the segmentation information, we continue to estimate the number of people in each segment by carrying out the blob analysis and blob size optimization approach. This approach however, provide useful information for understanding crowd dynamics yet it lacks significant information for understanding crowd behavior. Therefore, in this thesis, the current crowd flow segmentation and counting approach is further extended in order to coup the challenges of crowd behavior understanding. The extension adopts optical flow for the identification of pedestrian movements, and it considers the analyzed video as a set of sequences. The latter are analyzed separately, producing tracklets that are then clustered to produce global trajectories, defining both sources and sinks, but also characterizing the movement of pedestrians in the scene.
In the second part of the thesis, We propose a novel approach for automatic detection of social groups of pedestrians in crowds by considering only start (source) and stop (sink) locations of pedestrian trajectories. We build an Association Matrix that captures the joint probability distribution of starts and stops locations of all pedestrian trajectories to all other pedestrian trajectories in the scene. Pedestrians exhibiting similar distribution are combining in a group, where as similarity among the distributions is measuread by KL Divergence We adopt bottom-up hierarchical clustering approach, which is three step processes. In first step, we treat all the individuals as independent clusters, In the second step, couples are detected and after pruning of bad couples, Adjacency matrix is generated. Later on, in step three, using the Adjacency Matrix, groups of couples, those have strong intergroup closeness (similarity) are merged into a larger group..
8
Bisagno, Niccolò. "On simulating and predicting pedestrian trajectories in a crowd". Doctoral thesis, Università degli studi di Trento, 2020. http://hdl.handle.net/11572/256722.
Pełny tekst źródłaStreszczenie:
Crowds of people are gathering at multiple venues, such as concerts, political rallies, as well as in commercial malls, or just simply walking on the streets. More and more people are flocking to live in urban areas, thus generating a lot of scenarios of crowds. As a consequence, there is an increasing demand for automatic tools that can analyze and predict the behavior of crowds to ensure safety.
Crowd motion analysis is a key feature in surveillance and monitoring applications, providing useful hints about potential threats to safety and security in urban and public spaces. It is well known that people gatherings are generally difficult to model, due to the diversity of the agents composing the crowd. Each individual is unique, being driven not only by the destination but also by personality traits and attitude.
The domain of crowd analysis has been widely investigated in the literature. However, crowd gatherings have sometimes resulted in dangerous scenarios in recent years, such as stampedes or during dangerous situations.
To take a step toward ensuring the safety of crowds, in this work we investigate two main research problems: we try to predict each person future position and we try to understand which are the key factors for simulating crowds. Predicting in advance how a mass of people will fare in a given space would help in ensuring the safety of public gatherings.
9
Bisagno, Niccolò. "On simulating and predicting pedestrian trajectories in a crowd". Doctoral thesis, Università degli studi di Trento, 2020. http://hdl.handle.net/11572/256722.
Pełny tekst źródłaStreszczenie:
Crowds of people are gathering at multiple venues, such as concerts, political rallies, as well as in commercial malls, or just simply walking on the streets. More and more people are flocking to live in urban areas, thus generating a lot of scenarios of crowds. As a consequence, there is an increasing demand for automatic tools that can analyze and predict the behavior of crowds to ensure safety.
Crowd motion analysis is a key feature in surveillance and monitoring applications, providing useful hints about potential threats to safety and security in urban and public spaces. It is well known that people gatherings are generally difficult to model, due to the diversity of the agents composing the crowd. Each individual is unique, being driven not only by the destination but also by personality traits and attitude.
The domain of crowd analysis has been widely investigated in the literature. However, crowd gatherings have sometimes resulted in dangerous scenarios in recent years, such as stampedes or during dangerous situations.
To take a step toward ensuring the safety of crowds, in this work we investigate two main research problems: we try to predict each person future position and we try to understand which are the key factors for simulating crowds. Predicting in advance how a mass of people will fare in a given space would help in ensuring the safety of public gatherings.
10
Zäll, Emma. "Modelling Pedestrian-Induced Vertical Vibrations of Footbridges". Thesis, Umeå universitet, Institutionen för fysik, 2015. http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-101831.
Pełny tekst źródłaStreszczenie:
A pedestrian crowd walking on a footbridge causes the footbridge to vibrate. These vibrations become an issue of serviceability and can give rise to discomfort for the pedestrians, whereby they should, to as large extent as possible, be prevented. Currently, there is a lack of reliable models to describe a dynamic load on a footbridge, due to a walking crowd. Therefore, there is a need for such models. Lately, a great amount of research has been carried out on the subject pedestrian-induced vibrations of footbridges, though most of it with focus on lateral vibrations. Conversely, this project has been performed aiming to accurately model pedestrian-induced vertical vibrations of a general footbridge. For that purpose, starting from an existing model, a somewhat improved model, comprising three sub-model, has been developed. The sub-models are: one model of the pedestrian crowd walking along the footbridge, one model describing the load from the pedestrian footstep and one model describing the interaction between the pedestrians and the footbridge. In order to get statistically reliable results, numerous simulations of the pedestrian-induced vertical vibrations of a specific footbridge have been performed, using the developed model. Averaging the results over the simulations, we could conclude that the model gives an average error of 7 %, compared to experimental data. The measured quantity giving these results was the absolute maximum value of the acceleration at the midpoint of the footbridge. The achieved dynamical response of the footbridge is qualitatively satisfying, while the quantitative error is larger than we hoped for, whereby we conclude that further improvement of the model is needed before we are able to accurately model pedestrian-induced vertical vibrations of footbridges.När en folksamling går över en gångbro uppstår vibrationer i gångbron. Dessa vibrationer påverkar brons användbarhet och kan ge upphov till obehagskänsla hos fotgängarna, vilket gör att vibrationerna i största möjliga utsträckning bör motverkas. I nuläget saknas pålitliga modeller för att beskriva den dynamiska last en gångbro utsätts för när en folksamling går över den. Således föreligger ett behov att utveckla en sådan modell. Under de senaste decennierna har mycket forskning utförts inom området människoinducerade vibrationer i gångbroar. Dock har merparten av denna forskning berört endast laterala vibrationer. Detta projekt däremot, har genomförts med syftet att, med ett noggrant resultat, modellera människoinducerade vertikala vibrationer i en generell gångbro. För att uppnå detta har jag utgått från en befintlig modell och från den utvecklat en ny modell bestående av tre delmodeller. De tre delmodellerna är: en modell som beskriver hur folksamlingen rör sig över gångbron, en modell som beskriver den kraft det mänskliga fotsteget uträttar på gångbron och en modell som beskriver interaktionen mellan fotgängarna och gångbron. För att uppnå statistiskt tillförlitliga resultat har modellen som utvecklats i detta projekt använts för att utföra åtskilliga simuleringar av människoinducerade vertikala vibrationer i en specifik gångbro. Om vi medelvärdesbildar resultaten över simuleringarna framgår det att modellen som utvecklats ger ett resultat som avviker med 7 % från experimentellt data. Detta gäller för den maximala accelerationen vid gångbrons mittpunkt. Den resulterande dynamiska responsen ser kvalitativt sett bra ut, medan den kvantitativa avvikelsen är större än vi hoppats på. Därför drar vi slutsatsen att vidare förbättringar av modellen behövs för att den ska kunna användas till att på ett noggrant sätt modellera människoinducerade vertikala vibrationer i gångbroar.
11
Haciomeroglu, Murat. "Populating virtual urban environments with crowds of pedestrians in real-time". Thesis, University of East Anglia, 2009. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.518360.
Pełny tekst źródła
12
Qiu, Fasheng. "A Framework for Group Modeling in Agent-Based Pedestrian Crowd Simulations". Digital Archive @ GSU, 2010. http://digitalarchive.gsu.edu/cs_diss/56.
Pełny tekst źródłaStreszczenie:
Pedestrian crowd simulation explores crowd behaviors in virtual environments. It is extensively studied in many areas, such as safety and civil engineering, transportation, social science, entertainment industry and so on. As a common phenomenon in pedestrian crowds, grouping can play important roles in crowd behaviors. To achieve more realistic simulations, it is important to support group modeling in crowd behaviors. Nevertheless, group modeling is still an open and challenging problem. The influence of groups on the dynamics of crowd movement has not been incorporated into most existing crowd models because of the complexity nature of social groups. This research develops a framework for group modeling in agent-based pedestrian crowd simulations. The framework includes multiple layers that support a systematic approach for modeling social groups in pedestrian crowd simulations. These layers include a simulation engine layer that provides efficient simulation engines to simulate the crowd model; a behavior-based agent modeling layers that supports developing agent models using the developed BehaviorSim simulation software; a group modeling layer that provides a well-defined way to model inter-group relationships and intra-group connections among pedestrian agents in a crowd; and finally a context modeling layer that allows users to incorporate various social and psychological models into the study of social groups in pedestrian crowd. Each layer utilizes the layer below it to fulfill its functionality, and together these layers provide an integrated framework for supporting group modeling in pedestrian crowd simulations. To our knowledge this work is the first one to focus on a systematic group modeling approach for pedestrian crowd simulations. This systematic modeling approach allows users to create social group simulation models in a well-defined way for studying the effect of social and psychological factors on crowd’s grouping behavior. To demonstrate the capability of the group modeling framework, we developed an application of dynamic grouping for pedestrian crowd simulations.
13
Nishinari, Katsuhiro, Satoshi Kokubo i Kazuhiro Yamamoto. "Simulation for pedestrian dynamics by real-coded cellular automata (RCA)". Elsevier, 2007. http://hdl.handle.net/2237/20045.
Pełny tekst źródła
14
Shbib, Reda. "Pedestrians counting and event detection in crowded environment". Thesis, University of Portsmouth, 2015. https://researchportal.port.ac.uk/portal/en/theses/pedestrians-counting-and-event-detection-in-crowded-environment(76554ebe-0c0f-4409-9905-79b2f0b62bac).html.
Pełny tekst źródłaStreszczenie:
Crowd density estimation and pedestrian counting are becoming an area of interest such as assessing the social effect and impact between small groups of people within a crowd. Still, existing experimental crowd analyses performed by operators are time consuming. Generally, human controllers are engaged to achieve this task, however, more and more, visual surveillance are becoming an essential need, it is a hard task to watch and study all recorded video due to the huge number of cameras being installed. Currently, image-processing field has attracted all academic and research to develop automatic counting and monitoring algorithms. In this thesis, some novel contributions in different fields are presented: pedestrian counting, event detection, and queue monitoring. Firstly, this thesis presents an original contribution in the pedestrian counting domain. In recent years, many of proposed counting techniques have used global features to estimate crowd density. In this thesis, a new approach has been introduced to replace global image features by the low level- features, which are specific to individuals and clusters within the crowd. Thus, the total number of pedestrians is the summation of all clusters, which construct the crowd. Experimental results through different datasets showed that low-level features have performed better than global features. In addition to the pedestrian counting, this thesis presents another contribution in the area of pedestrian flow monitoring through the developing of a virtual door algorithm, in which pedestrians are counted while they are passing through a proposed virtual count line. Important features have been extracted from the region of interest. Discriminant features are detected, and optical flow of these points are assembled .The proposed system assembles optical flow in the trajectory direction in a discrete group of extracted feature points. Finally, this thesis presents a novel technique for estimating queue parameters, such as number of entrance, leaving and the frequency, in order to obtain a clear picture about the queue traffic and flow. Therefore, in order to obtain these parameters, the proposed pedestrian counting and virtual door approach have been integrated together. Experimental results conducted demonstrate that the proposed system is strong to real-life environments.
15
Pop, Dănuţ Ovidiu. "Multi-task cross-modality deep learning for pedestrian risk estimation Multi-task deep learning for pedestrian detection, action recognition and time to cross prediction". Thesis, Normandie, 2019. http://www.theses.fr/2019NORMIR06.
Pełny tekst źródłaStreszczenie:
Cette thèse de doctorat est le résultat de mes travaux de recherche dans le domaine de l'apprentissage automatique, du traitement d'image et du transport intelligent pour résoudre le problème du système de protection des piétons (PPS) multi-tâches comprenant non seulement la classification, la détection et le suivi des piétons, mais aussi l'action des piétons- classification et prédiction des unités, et enfin estimation du risque piéton. De plus, notre système PPS utilise des approches originales d'apprentissage en profondeur inter-modalités. Le but de notre travail de recherche est de développer un composant de protection des piétons intelligent basé uniquement sur un système de vision stéréo unique utilisant une architecture d'apprentissage en profondeur cross-modalité optimale afin de classer l'action piétonne actuelle, de prédire leurs prochaines actions et enfin d'estimer le piéton risque au moment de traverser pour chaque piéton. Premièrement, nous étudions la composante de classification où nous avons analysé comment les représentations d'apprentissage d'une modalité permettraient de reconnaître d'autres modalités au sein de divers apprentissages profonds, un terme comme apprentissage multimodal. Deuxièmement, nous étudions comment l'apprentissage inter-modalité améliore la détection de l'action piétonne de bout en bout.Troisièmement, nous analysons la prédiction de l'action des piétons et l'estimation du temps à traverserThis PhD thesis is the result of my research work in the machine learning, image processing and intelligent transportation field for solving the problem of multi-task pedestrian protection system (PPS) including not only pedestrian classification, detection and tracking, but also pedestrian action-unit classification and prediction, and finally pedestrian risk estimation. Moreover, our PPS system uses original cross-modality deep learning approaches. The goal of our research work is to develop an intelligent pedestrian protection component-based only on single stereo vision system using an optimal cross-modality deep learning architecture in order to classify the current pedestrian action, predict their next actions and finally to estimate the pedestrian risk by the time to cross for each pedestrian. First, we investigate the classification component where we analyzed how learning representations from one modality would enable recognition for other modalities within various deep learning, which one term as cross-modality learning. Second, we study how the cross-modality learning improves an end-to-end the pedestrian action
16
Chen, Ming. "Characterization of Pedestrian Electromagnetic Scattering at 76-77GHz". The Ohio State University, 2013. http://rave.ohiolink.edu/etdc/view?acc_num=osu1385579499.
Pełny tekst źródła
17
MANENTI, LORENZA ALESSANDRA. "Agent-based proxemic dynamics: crowd and groups simulation". Doctoral thesis, Università degli Studi di Milano-Bicocca, 2013. http://hdl.handle.net/10281/42374.
Pełny tekst źródłaStreszczenie:
Considering the general overview on the Pedestrian Dynamics area, this thesis is focused in the area of pedestrian dynamics simulation, with the goal to study the phenomenon of groups as constitutive elements that compose a crowd, analyzing if their presence influences the dynamics of pedestrian flow and evaluating the impact of their contribution.
18
Sorrentino, Luigi. "Simulation and optimization of crowd dynamics using a multiscale model". Doctoral thesis, Universita degli studi di Salerno, 2012. http://hdl.handle.net/10556/318.
Pełny tekst źródłaStreszczenie:
2010 - 2011In the last decades, the modeling of crowd motion and pedestrian .ow has attracted the attention of applied mathematicians, because of an increasing num- ber of applications, in engineering and social sciences, dealing with this or similar complex systems, for design and optimization purposes. The crowd has caused many disasters, in the stadiums during some major sporting events as the "Hillsborough disaster" occurred on 15 April 1989 at Hills- borough, a football stadium, in She¢ eld, England, resulting in the deaths of 96 people, and 766 being injured that remains the deadliest stadium-related disaster in British history and one of the worst ever international football accidents. Other example is the "Heysel Stadium disaster" occurred on 29 May 1985 when escaping, fans were pressed against a wall in the Heysel Stadium in Brussels, Belgium, as a result of rioting before the start of the 1985 European Cup Final between Liv- erpool of England and Juventus of Italy. Thirty-nine Juventus fans died and 600 were injured. It is well know the case of the London Millennium Footbridge, that was closed the very day of its opening due to macroscopic lateral oscillations of the structure developing while pedestrians crossed the bridge. This phenomenon renewed the interest toward the investigation of these issues by means of mathe- matical modeling techniques. Other examples are emergency situations in crowded areas as airports or railway stations. In some cases, as the pedestrian disaster in Jamarat Bridge located in South Arabia, mathematical modeling and numerical simulation have already been successfully employed to study the dynamics of the .ow of pilgrims, so as to highlight critical circumstances under which crowd ac- cidents tend to occur and suggest counter-measures to improve the safety of the event. In the existing literature on mathematical modeling of human crowds we can distinguish two approaches: microscopic and macroscopic models. In model at microscopic scale pedestrians are described individually in their motion by ordinary di¤erential equations and problems are usually set in two-dimensional domains delimiting the walking area under consideration, with the presence of obstacles within the domain and a target. The basic modeling framework relies on classical Newtonian laws of point. The model at the macroscopic scale consists in using partial di¤erential equations, that is in describing the evolution in time and space of pedestrians supplemented by either suitable closure relations linking the velocity of the latter to their density or analogous balance law for the momentum. Again, typical guidelines in devising this kind of models are the concepts of preferred direction of motion and discomfort at high densities. In the framework of scalar conservation laws, a macroscopic onedimensional model has been proposed by Colombo and Rosini, resorting to some common ideas to vehicular tra¢ c modeling, with the speci.c aim of describing the transition from normal to panic conditions. Piccoli and Tosin propose to adopt a di¤erent macroscopic point of view, based on a measure-theoretical framework which has recently been introduced by Canuto et al. for coordination problems (rendez-vous) of multiagent systems. This approach consists in a discrete-time Eulerian macroscopic representation of the system via a family of measures which, pushed forward by some motion mappings, provide an estimate of the space occupancy by pedestrians at successive time steps. From the modeling point of view, this setting is particularly suitable to treat nonlocal interactions among pedestrians, obstacles, and wall boundary conditions. A microscopic approach is advantageous when one wants to model di¤erences among the individuals, random disturbances, or small environments. Moreover, it is the only reliable approach when one wants to track exactly the position of a few walkers. On the other hand, it may not be convenient to use a microscopic approach to model pedestrian .ow in large environments, due to the high com- putational e¤ort required. A macroscopic approach may be preferable to address optimization problems and analytical issues, as well as to handle experimental data. Nonetheless, despite the fact that self-organization phenomena are often visible only in large crowds, they are a consequence of strategical behaviors devel- oped by individual pedestrians. The two scales may reproduce the same features of the group behavior, thus providing a perfect matching between the results of the simulations for the micro- scopic and the macroscopic model in some test cases. This motivated the multiscale approach proposed by Cristiani, Piccoli and Tosin. Such an approach allows one to keep a macroscopic view without losing the right amount of .granularity,.which is crucial for the emergence of some self-organized patterns. Furthermore, the method allows one to introduce in a macroscopic (averaged) context some micro- scopic e¤ects, such as random disturbances or di¤erences among the individuals, in a fully justi.able manner from both the physical and the mathematical perspec- tive. In the model, microscopic and macroscopic scales coexist and continuously share information on the overall dynamics. More precisely, the microscopic part tracks the trajectories of single pedestrians and the macroscopic part the density of pedestrians using the same evolution equation duly interpreted in the sense of measures. In this respect, the two scales are indivisible. Starting from model of Cristiani, Piccoli and Tosin we have implemented algo- rithms to simulate the pedestrians motion toward a target to reach in a bounded area, with one or more obstacles inside. In this work di¤erent scenarios have been analyzed in order to .nd the obstacle con.guration which minimizes the pedes- trian average exit time. The optimization is achieved using to algorithms. The .rst one is based on the exhaustive exploration of all positions: the average exit time for all scenarios is computed and then the best one is chosen. The second algorithm is of steepest descent type according to which the obstacle con.guration corresponding to the minimum exit time is found using an iterative method. A variant has been introduced to the algorithm so to obtain a more e¢ cient proce- dure. The latter allows to .nd better solutions in few steps than other algorithms. Finally we performed other simulations with bounded domains like a classical .at with .ve rooms and two exits, comparing the results of three di¤erent scenario changing the positions of exit doors. [edited by author]
X n.s.
19
GORRINI, ANDREA. "Empirical studies and computational results of a proxemic - based model of pedestrian crowd dynamics". Doctoral thesis, Università degli Studi di Milano-Bicocca, 2014. http://hdl.handle.net/10281/50254.
Pełny tekst źródłaStreszczenie:
The thesis work is organized in two main parts. The first includes a review of the social science framework about crowd dynamics and proxemics, and the methodological approach. The second part consists of several empirical studies. A summary of the contents is provided as follows. Starting from the pioneering study of Gustave Le Bon (1897), the social science contributions about crowds are reviewed in Chapter 2 (Contagion-Transformation Theory, Elaborated Social Identity Model, Emergent Norm Theory, Affiliative Approach). Chapter 3 presents the proxemic theory, with reference to the notion of personal space and the group proxemic behavior in static and motion situations. Chapter 4 presents the methodological approach, as composed of: in vivo observation, in vitro experiments and in silico simulations. Chapter 5 proposed the results achieved by means of two observations performed at the Campus of the University of Milano-Bicocca (Italy) and the Vittorio Emanuele II gallery (Milan, Italy). Chapter 6 presents two experiments focused on the combined impact of turning path and grouping on pedestrian crowd dynamics and the size of pedestrian personal space. Chapter 7 presents a simulation campaign performed by using the platform MAKKSim. The results achieved have been compared with the collected empirical data for sake of model validation. The thesis ends with final remarks about the achieved results and future works towards the improvement of the computational model of MAKKSim.
20
Xi, Hui. "A DDDAS-Based Multi-Scale Framework for Pedestrian Behavior Modeling and Interactions with Drivers". Diss., The University of Arizona, 2013. http://hdl.handle.net/10150/306361.
Pełny tekst źródłaStreszczenie:
A multi-scale agent-based simulation framework is firstly proposed to analyze pedestrian delays at signalized crosswalks in large urban areas under different conditions. The aggregated-level model runs under normal conditions, where each crosswalk is represented as an agent. Pedestrian counts collected near crosswalks are utilized to derive the binary choice probability from a utility maximization model. The derived probability function is utilized based on the extended Adam's model to estimate an average pedestrian delay with corresponding traffic flow rate and traffic light control at each crosswalk. When abnormality is detected, the detailed-level model with each pedestrian as an agent is running in the affected subareas. Pedestrian decision-making under abnormal conditions, physical movement, and crowd congestion are considered in the detailed-level model. The detailed-level model contains two sub-level models: the tactical sub-level model for pedestrian route choice and the operational sub-level model for pedestrian physical interactions. The tactical sub-level model is based on Extended Decision Field Theory (EDFT) to represent the psychological preferences of pedestrians with respect to different route choice options during their deliberation process after evaluating current surroundings. At the operational sub-level model, physical interactions among pedestrians and consequent congestions are represented using a Cellular Automata model, in which pedestrians are allowed biased random-walking without back step towards their destination that has been given by the tactical sub-level model. In addition, Dynamic-Data-Driven Application Systems (DDDAS) architecture has been integrated with the proposed multi-scale simulation framework for an abnormality detection and appropriate fidelity selection (between the aggregate level and the detailed level models) during the simulation execution process. Various experiments have been conducted under varying conditions with the scenario of a Chicago Loop area to demonstrate the advantage of the proposed framework, balancing between computational efficiency and model accuracy. In addition to the signalized intersections, pedestrian crossing behavior under unsignalized conditions which has been recognized as a main reason for pedestrian-vehicle crashes has also been analyzed in this dissertation. To this end, an agent-based model is proposed to mimic pedestrian crossing behavior together with drivers' yielding behavior in the midblock crossing scenario. In particular, pedestrian-vehicle interaction is first modeled as a Two-player Pareto game which develops evaluation of strategies from two aspects, delay and risk, for each agent (i.e. pedestrian and driver). The evaluations are then used by Extended Decision Field Theory to mimic decision making of each agent based on his/her aggressiveness and physical capabilities. A base car-following algorithm from NGSIM is employed to represent vehicles' physical movement and execution of drivers' decisions. A midblock segment of a typical arterial in the Tucson area is adopted to illustrate the proposed model, and the model for the considered scenario has been implemented in AnyLogic® simulation software. Using the constructed simulation, experiments have been conducted to analyze different behaviors of pedestrians and drivers and the mutual impact upon each other, i.e. average pedestrian delay resulted from different crossing behaviors (aggressive vs. conservative), and average braking distance which is affected by driving aggressiveness and drivers' awareness of pedestrians. The results look interesting and are believed to be useful for improvement of pedestrians' safety during their midblock crossing. To the best of our knowledge, the proposed multi-scale modeling framework for pedestrians and drivers is one of the first efforts to estimate pedestrian delays in an urban area with adaptive resolution based on demand and accuracy requirement, as well as to address pedestrian-vehicle interactions under unsignalized conditions.
21
CORBETTA, ALESSANDRO. "Multiscale Crowd Dynamics: Physical Analysis, Modeling and Applications". Doctoral thesis, Politecnico di Torino, 2016. http://hdl.handle.net/11583/2659720.
Pełny tekst źródłaStreszczenie:
In this thesis we investigate the dynamics of pedestrian crowds in a fundamental and applied perspective. Envisioning a quantitative understanding we employ ad hoc large-scale experimental measurements as well as analytic and numerical models. Moreover, we analyze current regulations in matter of pedestrians structural actions (structural loads), in view of the need of guaranteeing pedestrian safety in serviceable built environments.
This work comes in three complementary parts, in which we adopt distinct perspectives and conceptually different tools, respectively from statistical physics, mathematical modeling and structural engineering. Chapter 1 introduces these perspectives and gives an outline of the thesis.
The statistical dynamics of individual pedestrians is the subject of Part I.
Although individual trajectories may appear random, once we analyze them in large ensembles we expect ``preferred'' behaviors to emerge. Thus, we envisage individual paths as fluctuations around such established routes. To investigate this aspect, we perform year-long 24/7 measurements of pedestrian trajectories in real-life conditions, which we analyze statistically and via Langevin-like models.
Two measurement locations have been considered: a corridor-shaped landing in the Metaforum building at Eindhoven University of Technology and the main walkway within Eindhoven Train Station.
The measurement technique we employ, based on overhead Microsoft \Kinect\ 3D-range sensors and on ad hoc tracking algorithms, is introduced in Chapter 2.
In Chapter 3 we describe the low density pedestrian flows in the Metaforum landing. In this location hundreds of thousands of high-resolution trajectories have been collected.
First, we discuss standard crowd-traffic descriptions based on average quantities such as fundamental diagrams. Then, thanks to our large dataset, we address the dynamics beyond average values via probability distributions of pedestrian positions and velocities.
Chapter 4 focuses on the dynamics of pedestrians crossing the landing alone, i.e. undisturbed by peers. The simple crossing dynamics is affected by stochastic fluctuations due to the variability of individuals' behavior as well as external factors. In the chapter we propose a quantitative Langevin-like model for these stochastic fluctuations, that we compare with the experimental data in terms of stationary velocity distributions and time correlation functions.
The avoidance regime which takes place when two pedestrians walk simultaneously in the landing and in opposite directions is addressed in Chapter 5. In this regime, the statistical features of pedestrian motion change from the undisturbed case (Chapter 4).
Here, we study the avoidance dynamics as a linear superposition of the undisturbed motion and an interaction force. First, we estimate average interaction force fields from the data. Then, we extend the Langevin model of Chapter 4 to reproduce statistics of the pair-wise interactions.
Finally, in Chapter 6, we discuss in brief the measurements collected at Eindhoven Train Station in view of future dense crowd analyses.
In Part II we zoom out from the perspective of individual pedestrians and we look at crowds, adopting a genuine mathematical modeling point of view. In this context a microscopic, i.e. particle-like, or a macroscopic, i.e. fluid-like, observation scale can be employed. In Chapter 7, we establish a general background of crowd dynamics modeling, which includes an introduction of the modeling framework by Cristiani, Piccoli and Tosin (CPT framework, in use in Chapters 8,9,11 and 12.
This framework is suitable to model systems governed by social interactions and stands on a first order measure-valued evolution equation. The use of measures is crucial in the following, as it enables a unified treatment of crowd flows at the microscopic and macroscopic scales.
Chapter 8 comprises a comparison of microscopic and macroscopic dynamics given via the CPT framework. In a Wasserstein space context, we wonder when these two dynamics are consistent as the number of agents involved grows. In this comparison we consider agents whose mass (in a measure sense) is independent on the size of the crowd.
In Chapter 9 we focus on the modeling of crowds moving across elongated geometries resembling footbridges. We address pedestrians' motion in a macroscopic perspective via the CPT framework. According to the framework, dynamics are prescribed as a linear superposition of two components: a desired velocity (that encodes the motion of pedestrians walking alone) and a social velocity (that weights the crowd mass via an interaction kernel to assess individual reactions to mutual presence). Footbridge-like geometries are simple scenarios in which, from phenomenological considerations, we are able to give these components a reasonable form and thus perform simulations.
In Part III we consider crowd flows on footbridges in relation to the way the safety of pedestrians is ensured by the current building practice and in terms of crowd-structure dynamics interaction. Chapter 10 addresses crowd-footbridge systems in terms of featured uncertainties. We provide a categorized review of the literature giving a synthetic comparison of uncertainties involved. In general, beside the uncertainties affecting the mechanical properties of the structure, the status of the crowd is itself uncertain. Taking inspiration from wind engineering, we approach the crowd dynamics through a separation of the approaching and the crossing traffic. Within the review, we consider how building regulations address the crowd load. On one hand, no uncertainty, nor variability, is considered on the crowd state, therefore the roughest possible model (constant load) is typically retained. On the other hand, we notice how a large dissent is present in the prescribed load values, suggesting a possible inadequacy in regulations.
Chapter 11 rises from the point made in Chapter 10. We propose a framework to deal with uncertainties related to the crowd traffic on footbridges. The framework addresses the pedestrian density, a major player in the determination of live loads. Following the previous categorization, the framework is a composition of different modeling blocks and it considers approaching and crossing traffic at different scales, respectively macroscopic and microscopic. The output is a probabilistic description of the spatial density of the crossing crowd.
In Chapter 12 we consider the dynamics of the human-structure system as a whole, targeting the vertical vibrations of slender footbridges excited by crowds of walking pedestrians. We combine the microscopic counterpart of the CPT modeling framework for the pedestrian dynamics with a simple single degree of freedom structural model to provide a modeling framework for the crowd-structure interaction. We realize the coupling modeling the dynamical forces exchanged by the structure and each pedestrian via per-pedestrian single degree of freedom vertical oscillators. We study how the active presence of pedestrians influences the structure dynamics in terms of vertical accelerations and varied effective damping.
A discussion chapter addressing independently the content of the parts and then commenting on them as a whole closes the thesis.
22
Castle, Christian James Edward. "Agent-Based modelling of pedestrian evacuation: study of London's King's Cross underground station". Thesis, University College London (University of London), 2007. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.498192.
Pełny tekst źródła
23
Pellicanò, Nicola. "Tackling pedestrian detection in large scenes with multiple views and representations". Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLS608/document.
Pełny tekst źródłaStreszczenie:
La détection et le suivi de piétons sont devenus des thèmes phares en recherche en Vision Artificielle, car ils sont impliqués dans de nombreuses applications. La détection de piétons dans des foules très denses est une extension naturelle de ce domaine de recherche, et l’intérêt croissant pour ce problème est lié aux évènements de grande envergure qui sont, de nos jours, des scenarios à risque d’un point de vue de la sûreté publique. Par ailleurs, les foules très denses soulèvent des problèmes inédits pour la tâche de détection. De par le fait que les caméras ont le champ de vision le plus grand possible pour couvrir au mieux la foule les têtes sont généralement très petites et non texturées. Dans ce manuscrit nous présentons un système complet pour traiter les problèmes de détection et de suivi en présence des difficultés spécifiques à ce contexte. Ce système utilise plusieurs caméras, pour gérer les problèmes de forte occultation. Nous proposons une méthode robuste pour l’estimation de la position relative entre plusieurs caméras dans le cas des environnements requérant une surveillance. Ces environnements soulèvent des problèmes comme la grande distance entre les caméras, le fort changement de perspective, et la pénurie d’information en commun. Nous avons alors proposé d’exploiter le flot vidéo pour effectuer la calibration, avec l’objectif d’obtenir une solution globale de bonne qualité. Nous proposons aussi une méthode non supervisée pour la détection des piétons avec plusieurs caméras, qui exploite la consistance visuelle des pixels à partir des différents points de vue, ce qui nous permet d’effectuer la projection de l’ensemble des détections sur le plan du sol, et donc de passer à un suivi 3D. Dans une troisième partie, nous revenons sur la détection supervisée des piétons dans chaque caméra indépendamment en vue de l’améliorer. L’objectif est alors d’effectuer la segmentation des piétons dans la scène en partant d’une labélisation imprécise des données d’apprentissage, avec des architectures de réseaux profonds. Comme dernière contribution, nous proposons un cadre formel original pour une fusion de données efficace dans des espaces 2D. L’objectif est d’effectuer la fusion entre différents capteurs (détecteurs supervisés en chaque caméra et détecteur non supervisé en multi-vues) sur le plan du sol, qui représente notre cadre de discernement. nous avons proposé une représentation efficace des hypothèses composées qui est invariante au changement de résolution de l’espace de recherche. Avec cette représentation, nous sommes capables de définir des opérateurs de base et des règles de combinaison efficaces pour combiner les fonctions de croyance. Enfin, notre approche de fusion de données a été évaluée à la fois au niveau spatial, c’est à dire en combinant des détecteurs de nature différente, et au niveau temporel, en faisant du suivi évidentiel de piétons sur de scènes à grande échelle dans des conditions de densité variablePedestrian detection and tracking have become important fields in Computer Vision research, due to their implications for many applications, e.g. surveillance, autonomous cars, robotics. Pedestrian detection in high density crowds is a natural extension of such research body. The ability to track each pedestrian independently in a dense crowd has multiple applications: study of human social behavior under high densities; detection of anomalies; large event infrastructure planning. On the other hand, high density crowds introduce novel problems to the detection task. First, clutter and occlusion problems are taken to the extreme, so that only heads are visible, and they are not easily separable from the moving background. Second, heads are usually small (they have a diameter of typically less than ten pixels) and with little or no textures. This comes out from two independent constraints, the need of one camera to have a field of view as high as possible, and the need of anonymization, i.e. the pedestrians must be not identifiable because of privacy concerns.In this work we develop a complete framework in order to handle the pedestrian detection and tracking problems under the presence of the novel difficulties that they introduce, by using multiple cameras, in order to implicitly handle the high occlusion issues.As a first contribution, we propose a robust method for camera pose estimation in surveillance environments. We handle problems as high distances between cameras, large perspective variations, and scarcity of matching information, by exploiting an entire video stream to perform the calibration, in such a way that it exhibits fast convergence to a good solution. Moreover, we are concerned not only with a global fitness of the solution, but also with reaching low local errors.As a second contribution, we propose an unsupervised multiple camera detection method which exploits the visual consistency of pixels between multiple views in order to estimate the presence of a pedestrian. After a fully automatic metric registration of the scene, one is capable of jointly estimating the presence of a pedestrian and its height, allowing for the projection of detections on a common ground plane, and thus allowing for 3D tracking, which can be much more robust with respect to image space based tracking.In the third part, we study different methods in order to perform supervised pedestrian detection on single views. Specifically, we aim to build a dense pedestrian segmentation of the scene starting from spatially imprecise labeling of data, i.e. heads centers instead of full head contours, since their extraction is unfeasible in a dense crowd. Most notably, deep architectures for semantic segmentation are studied and adapted to the problem of small head detection in cluttered environments.As last but not least contribution, we propose a novel framework in order to perform efficient information fusion in 2D spaces. The final aim is to perform multiple sensor fusion (supervised detectors on each view, and an unsupervised detector on multiple views) at ground plane level, that is, thus, our discernment frame. Since the space complexity of such discernment frame is very large, we propose an efficient compound hypothesis representation which has been shown to be invariant to the scale of the search space. Through such representation, we are capable of defining efficient basic operators and combination rules of Belief Function Theory. Furthermore, we propose a complementary graph based description of the relationships between compound hypotheses (i.e. intersections and inclusion), in order to perform efficient algorithms for, e.g. high level decision making.Finally, we demonstrate our information fusion approach both at a spatial level, i.e. between detectors of different natures, and at a temporal level, by performing evidential tracking of pedestrians on real large scale scenes in sparse and dense conditions
24
Al-nasur, Sadeq J. "New Models for Crowd Dynamics and Control". Diss., Virginia Tech, 2006. http://hdl.handle.net/10919/30107.
Pełny tekst źródłaStreszczenie:
In recent years, there has been an increasing interest in modeling crowd and evacuation dynamics. Pedestrian models are based on macroscopic or microscopic behavior. In this work, we are interested in developing models that can be used for evacuation control strategies. Hence, we use macroscopic modeling approach, where pedestrians are treated in an aggregate way and detailed interactions are overlooked. In this dissertation, we developed two-dimensional space crowd dynamic models to allow bi-directional low by modifying and enhancing various features of existing traffic and fluid dynamic models. In this work, four models based on continuum theory are developed, and conservation laws such as the continuity and momentum equations are used. The first model uses a single hyperbolic partial differential equation with a velocity-density relationship, while the other three models are systems of hyperbolic partial differential equations. For one of the system models presented, we show how it can be derived independently from a microscopic crowd model. The models are nonlinear, time-varying, hyperbolic partial differential equations, and the numerical simulation results given for the four macroscopic models were based on computational fluid dynamics schemes.
We also started an initial control design that synthesizes the feedback linearization method for the one-dimensional traffic flow problem applied directly on the distributed parameter system. In addition, we suggest and discuss the information technology requirements for an evacuation system.
This research was supported in part from the National Science Foundation through grant no. CMS-0428196 with Dr. S. C. Liu as the Program Director. This support is gratefully acknowledged. Any opinion, findings, and conclusions or recommendations expressed in this study are those of the writer and do not necessarily reflect the views of the National Science Foundation.Ph. D.
25
Silva, Felipe Feliciano Gomes da. "Vibrações induzidas por multidões: efeito nos movimentos corpóreos dos pedestres na direção transversal". Universidade Federal da Paraíba, 2016. http://tede.biblioteca.ufpb.br:8080/handle/tede/8979.
Pełny tekst źródłaStreszczenie:
Submitted by Maike Costa (maiksebas@gmail.com) on 2017-06-05T12:42:07Z
No. of bitstreams: 1
arquivototal.pdf: 3657172 bytes, checksum: 49a2dbaae14b810a2f42bd26d7f5f3db (MD5)Made available in DSpace on 2017-06-05T12:42:07Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 3657172 bytes, checksum: 49a2dbaae14b810a2f42bd26d7f5f3db (MD5) Previous issue date: 2016-08-15
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES
The analysis of the vibrations induced by crowds in the transverse direction has become a topic of increasing interest, because of the ocurrence of excessive transverse vibrations in certain structures due to pedestrian traffic. This work presents the analysis of movements of pedestrians in the transverse direction for crowd situations. Successive tests were carried out with persons instrumented, walking alone and in a crowd. Comparisons between measurements of these pedestrians walking alone and in the crowd provided the identification of the influence of the crowd in the transverse acceleration and displacements and rotation of the pelvis.
A análise de vibrações induzidas por multidões na direção transversal vem se tornando um tema cada vez de maior interesse devido a ocorrência de vibrações transversais excessivas em algumas estruturas destinas à circulação de pedestres. Este trabalho apresenta a análise de caminhadas/movimentos na direção transversal para situações de multidão. Para obtenção dos dados experimentais foram feitos testes sucessivos com pessoas instrumentadas, caminhando isoladamente e em meio a uma multidão com densidades diferentes. A análise comparativa dos sinais destes pedestres caminhando só e em multidão, possibilitou a identificação da influência causada por essa multidão na aceleração e deslocamento transversal e na rotação da pélvis.
26
Cabrero, Daniel Beatriz. "Automating crowd simulation: from parameter tuning to dynamic context-to-policy adaptation". Doctoral thesis, Universitat Pompeu Fabra, 2022. http://hdl.handle.net/10803/673251.
Pełny tekst źródłaStreszczenie:
Computer-generated crowds are becoming more and more frequent in films, video games and safety assessment applications. Many crowd simulation algorithms exist to address the needs of this diverse range of industries. Even though the underlying principles are similar, there are large differences between the resulting synthetic trajectories. Each algorithm has strengths and weaknesses that need to be weighted, and appropriate parameter values for them must be selected as well. These are not easy tasks and Machine Learning algorithms are often used to guide these decisions. In this work we study three of these tasks: parameter tuning, trajectory evaluation, and character policy selection and adaptation. Our results show the usefulness of the proposed methods to evaluate previously unseen synthetic trajectories to find appropriate parameter values for the algorithms without directly relying on real data. Moreover, by classifying the context of characters, we propose a policy adaptation strategy to improve crowd simulations.Les multituds simulades per ordinador són cada cop més habituals en cinema, vídeo jocs i en aplicacions relacionades amb la seguretat. Existeixen molts algoritmes per simular multituds per adreçar tal varietat d’indústries. Tot i que els principis subjacents són similars, hi ha diferències entre les simulacions resultants. Cada algoritme té avantatges i inconvenients que s’han de valorar, i, a més a més, cal trobar valors pels seus paràmetres. Aquestes no són tasques senzilles i, sovint, es fan servir algoritmes d’aprenentatge automàtic per guiar aquestes decisions. Estudiem tres d’aquestes tasques: donar valor als paràmetres, avaluar trajectòries, i adaptar les polítiques. Els resultats demostren la utilitat dels mètodes proposats per avaluar trajectòries noves per tal de trobar valors apropiats pels paràmetres dels algorismes sense fer servir dades reals directament. A més a més, proposem una estratègia per adaptar la política de cada agent a través del reconeixement del context, millorant les simulacions.
27
Kabalan, Bachar. "Dynamique des foules : modélisation du mouvement des piétons et forces associées engendrées". Thesis, Paris Est, 2016. http://www.theses.fr/2016PESC1126/document.
Pełny tekst źródłaStreszczenie:
Que ce soit dans une rue commerçante, un supermarché ou un aéroport, les phénomènes de foule sont incontournables et nous affecte au quotidien. Elle constitue un système complexe dont la dynamique collective, résultant des interactions individuelles, est difficile à appréhender et a toujours intrigué les scientifiques de différents domaines. Grâce au progrès technologique, il est aujourd'hui possible de modéliser les mouvements de foule et de les reproduire en simulation. Les simulations de mouvement de foule permettent aux chercheurs de plusieurs disciplines, comme les sciences sociales ou la biomécanique, de mieux étudier et comprendre les mouvements des piétons et leurs interactions. Quant aux sciences de la sécurité et du transport, ils y voient des applications concrètes comme le développement de modèles de foule capables de simuler l'évacuation d'un lieu public de moyenne ou de forte affluence, afin que les futures constructions ou aménagements publics puissent offrir une qualité de sécurité et de service optimale pour les usagers. Dans le cadre de cette thèse, nous avons travaillé sur le perfectionnement du modèle discret proposé et développé par l'équipe dynamique du laboratoire Navier. Dans ce modèle, les actions et les décisions de chaque piéton sont traitées individuellement. Trois aspects du modèle ont été traités dans cette thèse. Le premier concerne la navigation des piétons vers leurs destinations. Dans notre modèle, un piéton est représenté par une particule ayant une direction et une allure souhaitées. Cette direction est obtenue par la résolution d'une équation eikonale. La solution de cette équation permet d'obtenir un champ de vitesses qui attribue à chaque piéton, en fonction de sa position, une direction vers sa destination. La résolution de l'équation une fois ou à une période quelconque donne la stratégie du chemin le plus court ou le plus rapide respectivement. Les effets des deux stratégies sur la dynamique collective de la foule sont comparés. Le deuxième consiste à gérer le comportement des piétons. Après avoir choisi son chemin, un piéton doit interagir avec l'environnement (obstacles, topologie, ...) et les autres piétons. Nous avons réussi à intégrer trois types de comportement dans notre modèle: (i) la poussée en utilisant une approche originale, basée sur la théorie des collisions des corps rigides dans un cadre thermodynamique rigoureux, (ii) le passage agressif (forcer son chemin) modélisé par une force sociale répulsive et (iii) l'évitement ``normal'' en adoptant une approche cognitive basée sur deux heuristiques. Les performances des trois méthodes ont été comparées pour plusieurs critères. Le dernier aspect concerne la validation et la vérification du modèle. Nous avons réalisé une étude de sensibilité et validé le modèle qualitativement et quantitativement. À l'aide d'un plan d'expérience numérique nous avons réussi à identifier les paramètres d'entrée ayant les effets principaux sur les résultats du modèle. De plus, nous avons trouvé les différentes interactions entre ces paramètres. En ce qui concerne la validation qualitative, nous avons réussi à reproduire plusieurs phénomènes d'auto-organisation. Enfin, nous avons testé la capacité de notre modèle à reproduire des résultats expérimentaux issus de la littérature. Nous avons choisi le cas du goulot d'étranglement. Les résultats du modèle et ceux de l'expérience ont été comparés. Ce modèle de foule a également été appliqué à l'acheminement des piétons dans la gare de Noisy-Champs. L'objectif de cette application est d'estimer le temps de stationnement des trains dans la gareCrowds are present almost everywhere and affect several aspects of our lives. They are considered to be on of the most complex systems whose dynamics, resulting from individual interactions and giving rise to fascinating phenomena, is very difficult to understand and have always intrigued experts from various domains. The technological advancement, especially in computer performance, has allowed to model and simulate pedestrian movement. Research from different disciplines, such as social sciences and bio-mechanics, who are interested in studying crowd movement and pedestrian interactions were able to better examine and understand the dynamics of the crowd. Professionals from architects and transport planners to fire engineers and security advisors are also interested in crowd models that would help them to optimize the design and operation of a facility. In this thesis, we have worked on the imporvement of a discrete crowd model developed by the researchers from the dynamics group in Navier laboratory. In this model, the actions and decisions taken by each individual are treated. In its previous version, the model was used to simulate urgent evacuations. Three main aspects of the model were addressed in this thesis. The first one concerns pedestrian navigation towards a final destination. In our model, a pedestrian is represented by a disk having a willingness to head to a certain destination with a desired direction and a desired speed. A desired direction is attributed to each pedestrian, depending on his position from the exit, from a floor field that is obtained by solving the eikonal equation. Solving this equation a single time at the beginning of the simulation or several times at during the simulation allows us to obtain the shortest path or the fastest path strategy respectively. The influence of the two strategies on the collective dynamics of the crowds is compared. The second one consists of managing pedestrian-pedestrian interactions. After having chosen his/her direction according to one of the available strategies, a pedestrian is bound to interact with other pedestrians present on the chosen path. We have integrated three pedestrian behaviors in our model: (i) pushing by using an original approach based on the theory of rigid body collisions in a rigorous thermodynamics context, (ii) forcing one's way by introducing a social repulsive force and (iii) "normal" avoidance by using a cognitive approach based on two heuristics. The three methods are compared for different criteria. The last aspect is the validation and verification of the model. We have performed a sensibility study and validated the model qualitatively and quantitatively. Using a numerical experimental plan, we identified the input parameters that are the most statistically significant and estimated the effects of their interactions. Concerning qualitative validation, we showed that our model is able to reproduce several self-organization phenomena such as lane formation. Finally, our model was validated quantitatively for the case of a bottleneck. The experimental results are very close to the ones obtained from simulations. The model was also applied to pedestrian movement in the Noisy-Champs train station. The objective of the study was to estimate the train dwell time. The simulation results were similar to the observations
28
Walker, Ross. "Autonomous robot navigation through a crowded and dynamic environment : using a novel form of path planning to demonstrate consideration towards pedestrians and other robots". Thesis, University of Sheffield, 2017. http://etheses.whiterose.ac.uk/18833/.
Pełny tekst źródłaStreszczenie:
This thesis presents a novel path planning algorithm for robotic crowd navigation through a pedestrian environment. The robot is designed to negotiate its way through the crowd using considerate movements. Unlike many other path planning algorithms, which assume cooperation with other pedestrians, this algorithm is completely independent and requires only observation. A considerate navigation strategy has been developed in this thesis, which utilises consideration as an directs an autonomous mobile robot. Using simple methods of predicting pedestrian movements, as well as simple relative distance and trajectory measurements between the robot and pedestrians, the robot can navigate through a crowd without causing disruption to pedestrian trajectories. Dynamic pedestrian positions are predicted using uncertainty ellipses. A novel Voronoi diagram-visibility graph hybrid roadmap is implemented so that the path planner can exploit any available gaps in between pedestrians, and plan considerate paths. The aim of the considerate path planner is to have the robot behave in specific ways when moving through the crowd. By predicting pedestrian trajectories, the robot can avoid interfering with them. Following preferences to move behind pedestrians, when cutting across their trajectories; to move in the same direction of the crowd when possible; and to slow down in crowded areas, will prevent any interference to individual pedestrians, as well as preventing an increase in congestion to the crowd as a whole. The effectiveness of the considerate navigation strategy is evaluated using simulated pedestrians, multiple mobile robots loaded with the path planning algorithm, as well as a real-life pedestrian dataset. The algorithm will highlight its ability to move with the aforementioned consideration towards each individual dynamic agent.
29
Paganelli, Lorenzo. "Simulazione di evacuazione di folle in Alchemist: un modello di mappa mentale per pedoni cognitivi". Bachelor's thesis, Alma Mater Studiorum - Università di Bologna, 2020. http://amslaurea.unibo.it/20540/.
Pełny tekst źródłaStreszczenie:
La simulazione computerizzata di folle in movimento è stato un campo piuttosto attivo recentemente, con applicazioni che vanno dall’intrattenimento alla gestione della sicurezza in luoghi pubblici o privati. Le simulazioni di folle devono prendere in considerazione non solo gli aspetti fisici dell’ambiente e dei pedoni, ma anche i fattori psicologici e sociali che hanno un effetto sul movimento delle persone. La capacità di orientarsi all’interno di un ambiente è una caratteristica fondamentale dell'essere umano e in quanto tale è indispensabile per una simulazione realistica. Diversi modelli assumono che i pedoni abbiano conoscenza totale dell’ambiente, cioè che ne conoscano la topologia e le metriche per intero. Ciò può essere ammissibile quando il processo di navigazione è banale, ad esempio in una piazza, tuttavia in ambienti più complessi questa è una grezza approssimazione, in quanto difficilmente ciascun pedone ne possiede una conoscenza completa (specie se molte delle persone coinvolte visitano l’edificio per la prima volta). Lo scopo del lavoro corrente è modellare la rappresentazione mentale che ciascun pedone ha dell’ambiente circostante, spesso parziale e inaccurata, comunemente nota come mappa cognitiva, e il processo mentale in atto in ogni persona che usa tali informazioni per scegliere quale percorso o direzione seguire (cioè orientarsi). I modelli descritti sono poi implementati all’interno del simulatore Alchemist, l'approccio adottato è ad agenti.
Di particolare interesse per il lavoro corrente sono le simulazioni di evacuazioni di folle, per il loro valore nel prevenire situazioni disastrose durante la pianificazione di eventi o la progettazione di edifici. Detto ciò, i modelli presentati sono pensati per essere usati in qualsiasi tipo di simulazione.
30
Du, Plessis Dewald. "APPOLO - Towards integrated urban education in Pretoria : a multi-functional vertical primary school". Diss., University of Pretoria, 2010. http://hdl.handle.net/2263/29976.
Pełny tekst źródłaStreszczenie:
The Apollo Project investigates the recent establishment of numerous private educational institutions in the inner city of Pretoria. It identifies the need for adequate urban educational facilities and explores the use of existing buildings as schools. An existing educational cluster is identified at the eastern edge of the inner city, defined by Church, Du Toit and Pretorius Streets, and Nelson Mandela Drive. This city block and the ones surrounding it contain numerous primary, secondary and tertiary educational institutions in a predominant industrial/automotive precinct. An urban design framework is proposed for the precinct. It is envisioned that the precinct may be developed as a mixed-use urban educational campus. Within the existing city block and the urban framework proposal, the Apollo Centre, located on the corner of Church-and Du Toit Street, is selected for an adaptive re-use intervention. The proposed use is an urban primary school. The Apollo project investigates current pedagogical trends, which informed a concept that is largely defined by the idea of contextual learning within a vertical structure. Transparency and integration of education with the urban environment is at the core of the proposal. The traditional notion of horizontal education is explored in a vertical manner. The existing structure is analyzed and a position taken regarding the adaptive re-use process that informs the design. Precedent Studies include existing schools within the inner city of Pretoria as well as local and international schools. The process of converting the Apollo Centre into a primary educational facility, that shares its resources on a cross-programming basis, is explored in a series of proposals. The numerous explorations are considered in their various aspects, as well as their relationship to the whole, which then leads to a final design proposal. Key areas of the proposed Apollo Primary School will finally be resolved technically. A conclusion summarizes the author’s thoughts on the result of the project.Mini Dissertation (MArch(Prof))--University of Pretoria, 2010.
Architecture
unrestricted
31
Johansson, Anders. "Data-Driven Modeling of Pedestrian Crowds". Doctoral thesis, 2008. https://tud.qucosa.de/id/qucosa%3A25055.
Pełny tekst źródłaStreszczenie:
At the starting point of the work leading to this doctoral thesis, in January 2005, the work on pedestrians was almost exclusively oriented towards computer simulations and on evacuation experiments. Since then, there have been many studies on new methods for extracting empirical data of pedestrian movements (mainly based on video analysis, lasers, and infrared cameras), but most of the work is still focused on artificial setups for crowds moving through corridors and crowds passing bottlenecks. Even though these controlled experiments are important to understand crowd dynamics, there is a knowledge gap between these experiments and the understanding of the dynamics leading to and occurring during large crowd disasters,
when sometimes hundreds of thousands or even millions of pedestrians are involved.
Numerous crowd disasters occur every year at large gatherings around the world. Unfortunately, the information
about the (spatio-temporal) development of these events tend to be qualitative rather than quantitative.
Video recordings from the crowd disaster
in Mina, Kingdom of Saudi Arabia, on the 12th of January 2006,
where hundreds of pilgrims lost their lives during the annual Muslim pilgrimage to Makkah,
gave the possibility to
scientifically evaluate the dynamics of the crowd.
With this video material, it was possible to observe and analyze the
behavior of the crowd under increasing crowd density, leading to the disaster.
Based on the insights from the analysis of the crowd disaster described above, new tools and measures
to detect and avoid critical crowd conditions
have been proposed, and some of them have been implemented in order to
reduce the likelihood of similar disasters in the future.
Further contributions of this thesis are to empirically evaluate many of the previous assumptions used for pedestrian
simulations. These assumptions are:
* A pedestrian avoids collisions by changing her or his walking speed by an acceleration which is exponentially decaying with the distance to the pedestrian or object being avoided.
* A pedestrian reacts stronger to what happens in front of her or him, than to what happens behind the back.
* The movement of a crowd of pedestrians always follows a smooth flow-density relationship, called the fundamental diagram.
* The walking speed will settle at 0 m/s at a specific maximum crowd density.
The first two assumptions were found to be consistent with the data,
but the pedestrian-flow theory had to be revised, since the two
latter assumptions do not always hold.
When these fundamental parts of pedestrian motion and avoiding maneuvers had been investigated,
an improved version of the social-force-model was formulated.
In order to enable the revision of previous works and the analysis of the crowd disaster mentioned above,
algorithms used for video-tracking have been introduced. The novelty of this work concerns the uniqueness and quantity of data on which the algorithms are validated and calibrated, but also
the focus on analyzing millions of pedestrians rather than hundreds.
The aim of this thesis is to move from theoretical models and controlled lab conditions to applicable models
for real-world conditions.Als diese Dissertation im Januar 2005 begonnen wurde, nutzten wissenschaftliche Untersuchungen von Fußgängern fast ausschließlich Computersimulationen und Evakuierungsexperimente. Seit dem haben viele Wissenschaftler an einer Verbesserung der Methoden gearbeitet. Heute werden empirische Daten mit Hilfe von Videoanalysen, Laser- und Infrarotkameras erhoben.Jedoch konzentrieren sich viele dieser Arbeiten auf künstliche Setups, in denen sich Fußgängermassen durch Korridore und Engpässe bewegen. Diese Experimente erlauben es, Massenbewegungen zu verstehen. Jedoch gibt es immer noch Forschungslücken. Es ist schwierig, unter solch kontrollierten Bedingungen Fortschritte darin zu erzielen, die auftretenden Dyamiken bei großen Katastrophen zu verstehen, in denen manchmal Hunderttausende oder sogar Millionen von Fußgängern involviert sind. Immer wieder kommt es zu Katastrophen in großen Menschenmengen. Leider sind von diesen Ereignissen häufig nur qualitative Informationen anstelle von quantitativen Daten erhältlich. Es ergab sich die besondere Gelegenheit, quantitatives Filmmaterial über eine Katastrophe in Mina (Königreich Saudi--Arabien) zu erhalten. Dort starben am 12. Januar 2006 hunderte von Pilgern während der jährlichen muslimischen Pilgerfahrt nach Mekka. Mit dem erhobenen Videomaterial konnte nachvollzogen werden, wie die Menschenmenge zuerst unbehindert fließen konnte, dann immer dichter wurde und wie es schließlich zur Katastrophe kam. Von den Erkenntnissen der Analyse der oben beschriebenen Katastrophe konnten neue Methoden entwickelt werden, die dabei helfen können, ähnliche Katastrophen in Zukunft zu vermeiden. Ein weiterer Beitrag dieser Dissertation besteht darin, einige Annahmen, die üblicherweise bei der Simulation von Fußgängerdynamiken gemacht werden, in Frage zu stellen und zu überarbeiten. Diese Annahmen sind: (1) Ein Fußgänger verhindert Zusammenstöße, indem er seine Schrittgeschwindigkeit so verändert, dass seine Beschleunigung exponentiell mit der Distanz zu dem zu umgehenden Fußgänger oder Objekt abnimmt. (2) Ein Fußgänger zeigt stärkere Reaktionen auf Ereignisse, die vor ihm passieren, als auf Ereignisse, die hinter ihm passieren. (3) Die Bewegung eines in einer Menschenmenge befindlichen Fußgängers folgt immer dem Strömungs--Dichte Verhältnis, was als Fundamental-Diagramm bezeichnet wird. (4) Die Laufgeschwindigkeit eines Fußgängers erreicht bei maximaler Menschendichte einem Wert von 0 m/s. Die ersten beiden Annahmen wurden von den empirischen Daten bestätigt. Unsere Analysen zeigen jedoch, dass die Annahmen 3 und 4 nicht immer gültig sind. Somit müssen Standardtheorien von Fußgängerdynamiken überarbeitet werden. Im Anschluß an die Analyse dieser fundamentalen Aspekte von Fußgängerverhalten und dem Verhalten bei Ausweichmanövern wird das Social-Force-Modell weiterentwickelt. Um auf vorhergehenden Arbeiten aufzubauen und um die oben beschriebene Katastrophe analysieren zu können, werden Algorithmen für die Video-Verfolgung von Fußgängerbewegungen entwickelt. Das Neue bei diesem Teil der Arbeit liegt nicht nur in dem verwendeten Verfahren selbst, sondern auch in der Einzigartigkeit und der großen Menge an verwendeten Daten, die mit diesem Verfahren analysiert werden. Ein zentrales Ziel dieser Arbeit besteht demnach in einer wissenschaftlichen Weiterentwicklung von theoretischen Modellen und kontrollierten Laborexperimenten hin zu Modellen, die unter realen Bedingungen tatsächlich anwendbar sind. Die Analyse von Fußgängern ist ein interdisziplinäres Feld, das von verschiedenen wissenschaftlichen Disziplinen mit verschiedenen Zielen betrieben wird. Leider gab es bislang wenig Bemühungen, die Resultate innerhalb dieser Teilgebiete im Rahmen einer konsistenten Theorie zu vereinen. Als seltene Ausnahmen können die Arbeiten von Teknomo und Antonini genannt werden. Diese Dissertation verfolgt das Ziel, diese theoretische Vereinigung weiter voran zu treiben. Dazu muss man zwischen der Neuerfindung des Rades und der Wiederverwendung nicht geprüfter Resultate abwägen. Dementsprechend ist ein Teil dieser Dissertation dem Vorhaben gewidmet, bisherige Forschung im Lichte empirischer Daten und neuer Methoden zu evaluieren. Da sich die Arbeit mit recht unterschiedlichen Aspekten von Fußgängerverhalten beschäftigt, konzentriert sich die Analyse in verschiedenen Teilen der Arbeit auf einige ausgewählte, alternative Modelle. Insbesondere bei der Modellierung und Simulation wird anstelle einer eingehenden Übersicht verschiedener Modelle eine Diskussion des speziellen Social-Force Modells präsentiert.
32
"Pedestrian Behavior Modeling and Understanding in Crowds". 2016. http://repository.lib.cuhk.edu.hk/en/item/cuhk-1292603.
Pełny tekst źródła
33
Chen, Shiang-yu, i 陳相宇. "Simulating Pedestrian Crowds in Smoky Emergency Situations". Thesis, 2008. http://ndltd.ncl.edu.tw/handle/47401495338216617404.
Pełny tekst źródłaStreszczenie:
碩士大同大學
資訊工程學系(所)
96
Crowd animation is often appearance in recent movie scenes. It can be realistic. Simulating behaviors and motion of people in emergency situations for safety systems has been widely studied recently. In this simulation, we simulate the crowd of pedestrians. Behaviors of people in emergency can be simulated by the system. When an architect designs the safety systems, how to set the safety equipment and, how to design the path of escaping from panic is an important problem. This system is based on a well known physics-based animation model which allows to consider influence of gaseous phenomena such as smoke in the behavior of the crowd. We simulate a normal public place, with the safety be set at the exits and paths. We can observe the effect of the smoke to the guiding lights received by pedestrians in this simulation.
34
Johansson, Anders Fredrik [Verfasser]. "Data-driven modeling of pedestrian crowds / by Anders Fredrik Johansson". 2009. http://d-nb.info/1007304812/34.
Pełny tekst źródła
35
Alrashed, Mohammed. "Agent Based Modeling and Simulation of Pedestrian Crowds In Panic Situations". Thesis, 2016. http://hdl.handle.net/10754/621941.
Pełny tekst źródłaStreszczenie:
The increasing occurrence of panic stampedes during mass events has motivated studying the impact of panic on crowd dynamics and the simulation of pedestrian flows in panic situations. The lack of understanding of panic stampedes still causes hundreds of fatalities each year, not to mention the scarce methodical studies of panic behavior capable of envisaging such crowd dynamics. Under those circumstances, there are thousands of fatalities and twice that many of injuries every year caused be crowd stampede worldwide, despite the tremendous efforts of crowd control and massive numbers of safekeeping forces. Pedestrian crowd dynamics are generally predictable in high-density crowds where pedestrians cannot move freely and thus gives rise to self-propelling interactions between pedestrians. Although every pedestrian has personal preferences, the motion dynamics can be modeled as a social force in such crowds. These forces are representations of internal preferences and objectives to perform certain actions or movements. The corresponding forces can be controlled for each individual to represent a different variety of behaviors that can be associated with panic situations such as escaping danger, clustering, and pushing. In this thesis, we use an agent-based model of pedestrian behavior in panic situations to predict the collective human behavior in such crowd dynamics. The proposed simulations suggests a practical way to alleviate fatalities and minimize the evacuation time in panic situations. Moreover, we introduce contagious panic and pushing behavior, resulting in a more realistic crowd dynamics model. The proposed methodology describes the intensity and spread of panic for each individual as a function of distances between pedestrians.
36
Khandelwal, Tarun. "Activity-travel behavior modeling of pilgrims in mass religious gatherings". Thesis, 2021. https://etd.iisc.ac.in/handle/2005/5660.
Pełny tekst źródłaStreszczenie:
The number of participants in mass gatherings like Kumbh Mela is ever increasing. Simulations for pre-event crowd modeling, risk assessment, and control planning can help set up robust crowd management and control mechanisms. However, it is necessary to understand better the processes and crowd movement patterns in mass gatherings to model and simulate crowds in such contexts. To this end, the activity-based modeling approach helps analyze the factors that influence different aspects of activity participation and time allocation for pilgrims. These aspects include the type of activities performed, the location and timing preferences for performing these activities, the time spent in different activity locations, etc. A good understanding of these factors can help in better modeling and simulating the spatio-temporal evolution of population density at the location of a mass gathering.
This thesis analyzes the factors influencing activity participation and duration for pilgrim groups and presents corresponding behavioral interpretations and policy implications. The groups mainly comprise non-resident individuals who are not necessarily from the same household. We use the activity diary and demographic data collected in the Kumbh Mela held in Ujjain in 2016 for our analyses. We contribute in the following ways towards the literature through our approach to analyze group behavior in mass religious gatherings: first, we abstract the activities into religious activities at ghat, temple, and camp locations, and identify the group demographic attributes that can influence the activity participation and duration. We then develop empirical models to analyze group activity participation and duration using binary logit, linear regression, and multiple discrete-continuous extreme value (MDCEV) models. The behavioral interpretations, backed up by findings from prior studies, and policy implications, appear consistent across our models. Our MDCEV models may be useful for aggregate activity time allocation along with models for location-specific allocations.
37
Lin, You-Kang, i 林祐綱. "Counting Pedestrians in Crowds under Occlusions". Thesis, 2009. http://ndltd.ncl.edu.tw/handle/71675718811723575015.
Pełny tekst źródłaStreszczenie:
碩士元智大學
電機工程學系
97
It’s difficult to segment and count people in occlusion. This problem has been an important task for a long time. Usually, we propose some hypotheses to simplify the problem of counting people in crowded. For example, the camera must be set the higher place top of humans. Therefore, the occlusion event between people will be simplified to blob merge and split. The problem of occlusion which far objects cover with near objects will be not produced. On the other hand, we can use multi cameras to surveillance the area which has overlap region between cameras. Then, use different information of cameras to make up for the problem of information is not enough in single camera. But those supposes are not suitable in real work. So, this thesis will count people under normal camera (one camera in side-view). This thesis uses contour matching to complete people counting quickly. It also uses markov chain monte carlo method to do the follow-up procedure of template matching. Finally, we use the result of template matching to create new contour model on-line.
38
Haji, Ali Abdul Lateef. "Pedestrian Flow in the Mean Field Limit". Thesis, 2012. http://hdl.handle.net/10754/250912.
Pełny tekst źródłaStreszczenie:
We study the mean-field limit of a particle-based system modeling the behavior of many indistinguishable pedestrians as their number increases. The base model is a modified version of Helbing's social force model. In the mean-field limit, the time-dependent density of two-dimensional pedestrians satisfies a four-dimensional integro-differential Fokker-Planck equation. To approximate the solution of the Fokker-Planck equation we use a time-splitting approach and solve the diffusion part using a Crank-Nicholson method. The advection part is solved using a Lax-Wendroff-Leveque method or an upwind Backward Euler method depending on the advection speed.
Moreover, we use multilevel Monte Carlo to estimate observables from the particle-based system. We discuss these numerical methods, and present numerical results showing the convergence of observables that were calculated using the particle-based model as the number of pedestrians increases to those calculated using the probability density function satisfying the Fokker-Planck equation.
39
YU, SHU-HAN, i 虞宗翰. "Pedestrian Re-Identification and Crowd Tracking Using Convolutional Neural Networks". Thesis, 2019. http://ndltd.ncl.edu.tw/handle/9972dw.
Pełny tekst źródłaStreszczenie:
碩士國立臺北大學
資訊工程學系
107
Plenty of surveillance cameras were installed in several public areas,the camera video stream control monitoring system can assist in maintaining safety at the site to prevent criminal activities . Without a surveillance camera system, a large number of human patrols or guards must be arranged to achieve the same level of monitoring, which will increase personnel and logistical costs. This study proposes a method for pedestrian re-identification and crowd tracking, using computer vision techniques to detect targets from different camera views and angles, as well as identifying and tracking pedestrians or populations and calculating the number of people. We use Mask R-CNN to detect pedestrians. Based on the results of the detection, tracking is performed using the IOU-tracker. The tracking results are compared using a batch-feature- erasing-network to confirm the similarity between data from different cameras. The crowd is defined by the results of re-identification, after which it is tracked and counted, and finally, re-identification and tracking is fully achieved using this method. The method shows good results with the Duke dataset.
40
Chen, Chun-Yuan, i 陳俊元. "Crowded Pedestrian Detection Using EM based on Weighted Local Features". Thesis, 2014. http://ndltd.ncl.edu.tw/handle/49492808863148476349.
Pełny tekst źródłaStreszczenie:
碩士國立高雄第一科技大學
電腦與通訊工程研究所
102
Pedestrian detection and counting is an important topic in developing an intelligent surveillance system. In this work, we propose a vision-based system for detecting pedestrians in an image. Be robust to crowded scenes and adapt to incomplete foreground from background subtraction algorithm, expectation maximization (EM) algorithm is applied to impose the constraint of body part for achieving successful detection. First, the corner points at body part are all detected and described using histogram of oriented gradients (HOGs). In addition, one of three body part labels (head, torso, and leg), a kind of locality property, is encoded in corner points for overcoming the mutual occlusion situation. Then, we apply a grouping algorithm in HOGs feature space to form a set of clusters. Each cluster center is considered as a code word and the probabilities of this cluster belonging to head, torso, or leg are also computed, respectively. During detecting phase, all detected corner points are matched to the construct code words and are assigned to three body part probabilities. After that, an EM algorithm is applied to iteratively estimate the likelihood probability of all corner points to the pedestrian candidates (E-Step) and update the parameters of the pedestrian models (M-Step). In the experiment, three videos are used to validate the performance of the proposed method.
41
Chen, Yao-Hsiang, i 陳耀祥. "Crowded Pedestrian Detection Using EM based on Multi-level HOGs of Body Parts". Thesis, 2015. http://ndltd.ncl.edu.tw/handle/25720267977881339799.
Pełny tekst źródłaStreszczenie:
碩士國立高雄第一科技大學
電腦與通訊工程研究所
103
This paper proposes a vision-based method for detecting pedestrians from crowded scenes. To address occlusion problem faced in crowded pedestrian detection, this paper models the pedestrian appearance by equally dividing a pedestrian into eight body patches. Each patch is described by using multi-level histogram of oriented gradients (ML-HOGs) and then a support vector machine (SVM) is learned for determining its existence. By parameterizing the pedestrians to be detected, the problem of crowded pedestrian detection is formulated as the parameter estimation one. A statistical way to solve this problem is by maximizing likelihood (ML) probability. A well-known algorithm called expectation maximization (EM) is applied to achieve the solution in case of missing data due to occlusion in crowded scenes. The EM algorithm firstly generates a set of initial pedestrian hypothesis by using the learned eight SVMs to detect the patches with high possibility belonging to the head through voting mechanism. After that, EM algorithm iteratively estimates the likelihood probability of all pixels with respect to the pedestrian hypothesis (E-step) and then updates their parameters (M-step) until it reaches the convergence condition. Finally, the false and redundant hypothesis from EM algorithm are removed through the proposed verification steps. In the experiment, the videos from CAVIAR datasets are used for validating the proposed method.
42
Yaghoubi, Ehsan. "Soft Biometric Analysis: MultiPerson and RealTime Pedestrian Attribute Recognition in Crowded Urban Environments". Doctoral thesis, 2021. http://hdl.handle.net/10400.6/12081.
Pełny tekst źródłaStreszczenie:
Traditionally, recognition systems were only based on human hard biometrics. However,
the ubiquitous CCTV cameras have raised the desire to analyze human biometrics from
far distances, without people attendance in the acquisition process. Highresolution
face closeshots
are rarely available at far distances such that facebased
systems cannot
provide reliable results in surveillance applications. Human soft biometrics such as body
and clothing attributes are believed to be more effective in analyzing human data collected
by security cameras.
This thesis contributes to the human soft biometric analysis in uncontrolled environments
and mainly focuses on two tasks: Pedestrian Attribute Recognition (PAR) and person reidentification
(reid).
We first review the literature of both tasks and highlight the history
of advancements, recent developments, and the existing benchmarks. PAR and person reid
difficulties are due to significant distances between intraclass
samples, which originate
from variations in several factors such as body pose, illumination, background, occlusion,
and data resolution. Recent stateoftheart
approaches present endtoend
models that
can extract discriminative and comprehensive feature representations from people. The
correlation between different regions of the body and dealing with limited learning data
is also the objective of many recent works. Moreover, class imbalance and correlation
between human attributes are specific challenges associated with the PAR problem.
We collect a large surveillance dataset to train a novel gender recognition model suitable
for uncontrolled environments. We propose a deep residual network that extracts several
posewise
patches from samples and obtains a comprehensive feature representation. In
the next step, we develop a model for multiple attribute recognition at once. Considering
the correlation between human semantic attributes and class imbalance, we respectively
use a multitask
model and a weighted loss function. We also propose a multiplication
layer on top of the backbone features extraction layers to exclude the background features
from the final representation of samples and draw the attention of the model to the
foreground area.
We address the problem of person reid
by implicitly defining the receptive fields of
deep learning classification frameworks. The receptive fields of deep learning models
determine the most significant regions of the input data for providing correct decisions.
Therefore, we synthesize a set of learning data in which the destructive regions (e.g.,
background) in each pair of instances are interchanged. A segmentation module
determines destructive and useful regions in each sample, and the label of synthesized
instances are inherited from the sample that shared the useful regions in the synthesized
image. The synthesized learning data are then used in the learning phase and help
the model rapidly learn that the identity and background regions are not correlated.
Meanwhile, the proposed solution could be seen as a data augmentation approach that
fully preserves the label information and is compatible with other data augmentation
techniques.
When reid
methods are learned in scenarios where the target person appears with identical garments in the gallery, the visual appearance of clothes is given the most
importance in the final feature representation. Clothbased
representations are not
reliable in the longterm
reid
settings as people may change their clothes. Therefore,
developing solutions that ignore clothing cues and focus on identityrelevant
features are
in demand. We transform the original data such that the identityrelevant
information of
people (e.g., face and body shape) are removed, while the identityunrelated
cues (i.e.,
color and texture of clothes) remain unchanged. A learned model on the synthesized
dataset predicts the identityunrelated
cues (shortterm
features). Therefore, we train a
second model coupled with the first model and learns the embeddings of the original data
such that the similarity between the embeddings of the original and synthesized data is
minimized. This way, the second model predicts based on the identityrelated
(longterm)
representation of people.
To evaluate the performance of the proposed models, we use PAR and person reid
datasets, namely BIODI, PETA, RAP, Market1501,
MSMTV2,
PRCC, LTCC, and MIT
and compared our experimental results with stateoftheart
methods in the field.
In conclusion, the data collected from surveillance cameras have low resolution, such
that the extraction of hard biometric features is not possible, and facebased
approaches
produce poor results. In contrast, soft biometrics are robust to variations in data quality.
So, we propose approaches both for PAR and person reid
to learn discriminative features
from each instance and evaluate our proposed solutions on several publicly available
benchmarks.This thesis was prepared at the University of Beria Interior, IT Instituto de Telecomunicações, Soft Computing and Image Analysis Laboratory (SOCIA Lab), Covilhã Delegation, and was submitted to the University of Beira Interior for defense in a public examination session.
43
Liang, Chih-Bin, i 梁志彬. "A Vision-based Detection System for Walking-cross Activities of Pedestrians". Thesis, 1997. http://ndltd.ncl.edu.tw/handle/18872406598533197589.
Pełny tekst źródłaStreszczenie:
碩士淡江大學
土木工程學系
85
In this thesis, we try to use image processing technique to count pedestrians which are stationary in the image and track their trajectories when they are walking. In the pedestrian counting system, since the pixels which pedestrian possesses in the image are directly proportional to number of pedestrians, so we use background differencing method to establish regression model and calculate all people in the image. On the other hand, we use interframe differencing method to establish another regression model which is used to calculate moving pedestrians in the image. The difference between total number of pedestrians and number of moving pedestrians at the same time is considered to be the number of stationary pedestrians in the image. About pedestrian''s trajectory detection, we adopt template matching method based on image feature invariants to track each pedestrian''s template in sequence images. After linking every coordinates, we obtain pedestrians'' trajectory lines. The detection system can output four important parameters including pedestrians'' number, trajectories, speeds, and directions. A case study was demonstrated. About stationary pedestrians, the absolute diffenence between automaticcounts and real counts for each image is almost within 2 persons, and the ratio of maximum absolute error to the real count decrease gradually when people increase. About trajectory, the use of template matching method have only an average error of 1 pixel in x-axis, 2 pixel in y-axis. The error of speed is about 9.43%.
44
Alrashed, Mohammed. "Control Theoretic Approaches to Computational Modeling and Risk Mitigation for Large Crowd Management". Diss., 2020. http://hdl.handle.net/10754/665965.
Pełny tekst źródłaStreszczenie:
We develop a computational framework for risk mitigation in high population density events. With increased global population, the frequency of high population density events is naturally increased. Therefore, risk-free crowd management plans are critical for efficient mobility, convenient daily life, resource management and most importantly mitigation of any inadvertent incidents and accidents such as stampedes. The status-quo for crowd management plans is the use of human experience/expert advice. However, most often such dependency on human experience is insufficient, flawed and results in inconvenience and tragic events. Motivated by these issues, we propose an agent-based mathematical model describing realistic human motion and simulating large dense crowds in a wide variety of events as a potential simulation testbed to trial crowd management plans. The developed model incorporates stylized mindset characteristics as an internal drive for physical behavior such as walking, running, and pushing. Furthermore, the model is combined with a visualisation of crowd movement. We develop analytic tools to quantify crowd dynamic features. The analytic tools will enable verification and validation to empirical evidence and surveillance video feed in both local and holistic representations of the crowd. This work addresses research problems in computational modeling of crowd dynamics, specifically: understanding and modeling the impact of a collective mindset on crowd dynamics versus mixtures of heterogeneous mindsets, the effect of social contagion of behaviors and decisions within the crowd, the competitive and aggressive pushing behaviors, and torso and steering dynamics.