Video surveillance has long been in use to monitor security sensitive areas such as banks, department stores, crowded public places and borders. The rise in computer speed, availability of cheap large-capacity storage devices and high speed network infrastructure enabled the way for cheaper, multi sensor video surveillance systems. In this thesis, the problem of tracking multiple targets with multiple cameras has been discussed. Cameras have been located so that they have overlapping fields of vision. A dynamic background-modeling algorithm is described for segmenting moving objects from the background, which is capable of adapting to dynamic scene changes and periodic motion, such as illumination change and swaying of trees. After segmentation of foreground scene, the objects to be tracked have been acquired by morphological operations and connected component analysis. For the purpose of tracking the moving objects, an active contour model (snakes) is one of the approaches, in addition to a Kalman tracker. As the main tracking algorithm, a rule based tracker has been developed first for a single camera, and then extended to multiple cameras. Results of used and proposed methods are given in detail.

Thesis (MScEng (Mathematical Sciences)--University of Stellenbosch, 2009. / ENGLISH ABSTRACT: To reliably classify parts of an image sequence as foreground or background is an important part of many computer vision systems, such as video surveillance, tracking and robotics. It can also be important in applications where bandwidth is the limiting factor, such as video conferencing. Independent foreground motion is an attractive source of information for this task, and with static cameras, background subtraction is a particularly popular type of approach. The idea behind background subtraction is to compare the current image with a reference image of the background, and from there decide on a pixel by pixel basis, what is foreground and what is background by observing the changes in the pixel sequence. The problem is to get the useful reference image, especially when large parts of the background are occluded by moving/stationary foreground objects; i.e. some parts of the background are never seen. In this thesis four algorithms are reviewed that segment an image sequence into foreground and background components with varying degrees of success that can be measured on speed, comparative accuracy and/or memory requirements. These measures can be then effectively used to decide the application scope of the individual algorithms. / AFRIKAANSE OPSOMMING: Om betroubaar dele van ’n beeld reeks te klassifiseer as voorgrond of agtergrond is ’n belangrike deel van baie rekenaarvisie sisteme, byvoorbeeld video bewaking, volging en robotika. Dit kan ook belangrik wees in toepassings waar bandwydte die beperkende faktor is, byvoorbeeld video konferensie gesprekke. Onafhanklik voorgrond beweging is ’n aantreklike bron van informasie vir hierdie taak, en met statiese kameras, is agtergrond aftrekking ’n populêre benadering. Die idee agter agtergrond aftrekking is om die huidige beeld met ’n naslaan beeld van die agtergrond te vergelyk, en daarvandaan besluit op ’n piksel-na-piksel basis, wat is voorgrond en wat is agtergrond deur die observasies van die veranderinge in die piksel-reeks. Die probleem is om die naslaan beeld te kry om mee te werk, veral wanneer groot dele van die agtergrond onsigbaar bly as gevolg van bewegende of stilstaande voorgrond objekte en sommige dele van die agtergrond word dalk nooit gesien nie. In hierdie tesis word vier algorithms ondersoek wat ’n beeld reeks segmenteer in respektiewe voorgrond en agtergrond komponente met wisselende grade van sukses wat gemeet kan word deur spoed, vergelykbare akkuraatheid en/of geheu gebruik. Hierdie metings kan dan effektief gebruik word om die applikasie veld van die individuele algoritmes the bepaal.

Submitted by JÚLIO HEBER SILVA (julioheber@yahoo.com.br) on 2017-06-23T19:10:20Z No. of bitstreams: 2 Dissertação - André Reis de Geus - 2016.pdf: 8269330 bytes, checksum: 1345e49235c545021c88a7baf696f5c0 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Cláudia Bueno (claudiamoura18@gmail.com) on 2017-07-07T20:25:28Z (GMT) No. of bitstreams: 2 Dissertação - André Reis de Geus - 2016.pdf: 8269330 bytes, checksum: 1345e49235c545021c88a7baf696f5c0 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-07-07T20:25:28Z (GMT). No. of bitstreams: 2 Dissertação - André Reis de Geus - 2016.pdf: 8269330 bytes, checksum: 1345e49235c545021c88a7baf696f5c0 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2016-03-10 / Fundação de Amparo à Pesquisa do Estado de Goiás - FAPEG / Insects cause significant losses of stored grains in both quantity and quality. In the scenary, it is of paramount importance an early identification of insects in grains to take control measures. Instead of sampling and visual/laboratory analysis of grains, we propose to carry out the insects identification task automatically, using computational methods to perform endoscopic video analysis. The videos are recorded inside of grains warehouses by an endoscopic camera. As the classification process of moving objects in video rely heavily on precise segmentation of moving objets, we propose a new method of background subtraction and compared their results with the main methods of the literature according to a recent review. The main innovation of the background subtractionmethod rely on the binarization process that uses two thresholds: a global and a local threshold. The binarized results are combined by adding details of the object obtained by the local threshold in the result of the global threshold. Experimental results performed through visual analysis of the segmentation results and using a SVM classifier, suggest that the proposed segmentation method produces more accurate results than the state-of-art background subtraction methods. / Insetos causam perdas quantitativas e qualitativas significantesemgrãos armazenados. Neste cenário, é de vital importância uma identificação rápida de insetos em grãos para que sejam tomadas medidas de controle. Ao invés de coletar amostras de grãos para análise visual/laboratorial, é proposta a realização desta tarefa de identificação de formaautomática, usando métodos computacionais para a análise de vídeos endoscópicos. Os vídeos são gravados dentro de armazéns de grãos usando câmera endoscópica. Como o processo de classificação de objetos em movimento em vídeo depende fundamentalmente de uma segmentação de objeto precisa, é proposto um novo método de segmentação por subtração de plano de fundo e comparado seus resultados com os principais métodos da literatura de acordo com um estudo de revisão recente. A principal inovação neste método de subtração de plano de fundo está no processo de binarização que usa dois thresholds: um global e um local. Os resultados binarizados são combinados pela adição de detalhes do objeto obtido pelo threshold local no resultado do threshold global. Resultados experimentais, realizados através de análise visual dos resultados de segmentação e usandoumclassificadorSVMindicamque o método de segmentação proposto produz melhores resultados que métodos do estado da arte atual da literatura de subtração de plano de fundo.

Foreground segmentation is a common first step in tracking and surveillance applications.  The purpose of foreground segmentation is to provide later stages of image processing with an indication of where interesting data can be found.  This thesis is an investigation of how foreground segmentation can be performed in two contexts: as a pre-step to trajectory tracking and as a pre-step in indoor surveillance applications. Three methods are selected and detailed: a single Gaussian method, a Gaussian mixture model method, and a codebook method.  Experiments are then performed on typical input video using the methods.  It is concluded that the Gaussian mixture model produces the output which yields the best trajectories when used as input to the trajectory tracker.  An extension is proposed to the Gaussian mixture model which reduces shadow, improving the performance of foreground segmentation in the surveillance context.

Le suivi d’objets est de plus en plus utilisé dans les applications de vision par ordinateur. Compte tenu des exigences des applications en termes de performance, du temps de traitement, de la consommation d’énergie et de la facilité du déploiement des systèmes de suivi, l’utilisation des architectures embarquées de calcul devient primordiale. Dans cette thèse, nous avons conçu un système de suivi d’objets pouvant fonctionner en temps réel sur une caméra intelligente de faible coût et de faible consommation équipée d’un processeur embarqué ayant une architecture légère en ressources de calcul. Le système a été étendu pour le suivi d’objets dans un réseau de caméras avec des champs de vision non-recouvrant. La chaîne algorithmique est composée d’un étage de détection basé sur la soustraction de fond et d’un étage de suivi utilisant un algorithme probabiliste Gaussian Mixture Probability Hypothesis Density (GMPHD). La méthode de soustraction de fond que nous avons proposée combine le résultat fournie par la méthode Zipfian Sigma-Delta avec l’information du gradient de l’image d’entrée dans le but d’assurer une bonne détection avec une faible complexité. Le résultat de soustraction est traité par un algorithme d’analyse des composantes connectées afin d’extraire les caractéristiques des objets en mouvement. Les caractéristiques constituent les observations d’une version améliorée du filtre GMPHD. En effet, le filtre GMPHD original ne traite pas les occultations se produisant entre les objets. Nous avons donc intégré deux modules dans le filtre GMPHD pour la gestion des occultations. Quand aucune occultation n’est détectée, les caractéristiques de mouvement des objets sont utilisées pour le suivi. Dans le cas d’une occultation, les caractéristiques d’apparence des objets, représentées par des histogrammes en niveau de gris sont sauvegardées et utilisées pour la ré-identification à la fin de l’occultation. Par la suite, la chaîne de suivi développée a été optimisée et implémentée sur une caméra intelligente embarquée composée de la carte Raspberry Pi version 1 et du module caméra RaspiCam. Les résultats obtenus montrent une qualité de suivi proche des méthodes de l’état de l’art et une cadence d’images de 15 − 30 fps sur la caméra intelligente selon la résolution des images. Dans la deuxième partie de la thèse, nous avons conçu un système distribué de suivi multi-objet pour un réseau de caméras avec des champs non recouvrants. Le système prend en considération que chaque caméra exécute un filtre GMPHD. Le système est basé sur une approche probabiliste qui modélise la correspondance entre les objets par une probabilité d’apparence et une probabilité spatio-temporelle. L’apparence d’un objet est représentée par un vecteur de m éléments qui peut être considéré comme un histogramme. La caractéristique spatio-temporelle est représentée par le temps de transition des objets et la probabilité de transition d’un objet d’une région d’entrée-sortie à une autre. Le temps de transition est modélisé par une loi normale dont la moyenne et la variance sont supposées être connues. L’aspect distribué de l’approche proposée assure un suivi avec peu de communication entre les noeuds du réseau. L’approche a été testée en simulation et sa complexité a été analysée. Les résultats obtenus sont prometteurs pour le fonctionnement de l’approche dans un réseau de caméras intelligentes réel. / Multi-object tracking constitutes a major step in several computer vision applications. The requirements of these applications in terms of performance, processing time, energy consumption and the ease of deployment of a visual tracking system, make the use of low power embedded platforms essential. In this thesis, we designed a multi-object tracking system that achieves real time processing on a low cost and a low power embedded smart camera. The tracking pipeline was extended to work in a network of cameras with nonoverlapping field of views. The tracking pipeline is composed of a detection module based on a background subtraction method and on a tracker using the probabilistic Gaussian Mixture Probability Hypothesis Density (GMPHD) filter. The background subtraction, we developed, is a combination of the segmentation resulted from the Zipfian Sigma-Delta method with the gradient of the input image. This combination allows reliable detection with low computing complexity. The output of the background subtraction is processed using a connected components analysis algorithm to extract the features of moving objects. The features are used as input to an improved version of GMPHD filter. Indeed, the original GMPHD do not manage occlusion problems. We integrated two new modules in GMPHD filter to handle occlusions between objects. If there are no occlusions, the motion feature of objects is used for tracking. When an occlusion is detected, the appearance features of the objects are saved to be used for re-identification at the end of the occlusion. The proposed tracking pipeline was optimized and implemented on an embedded smart camera composed of the Raspberry Pi version 1 board and the camera module RaspiCam. The results show that besides the low complexity of the pipeline, the tracking quality of our method is close to the stat of the art methods. A frame rate of 15 − 30 was achieved on the smart camera depending on the image resolution. In the second part of the thesis, we designed a distributed approach for multi-object tracking in a network of non-overlapping cameras. The approach was developed based on the fact that each camera in the network runs a GMPHD filter as a tracker. Our approach is based on a probabilistic formulation that models the correspondences between objects as an appearance probability and space-time probability. The appearance of an object is represented by a vector of m dimension, which can be considered as a histogram. The space-time features are represented by the transition time between two input-output regions in the network and the transition probability from a region to another. Transition time is modeled as a Gaussian distribution with known mean and covariance. The distributed aspect of the proposed approach allows a tracking over the network with few communications between the cameras. Several simulations were performed to validate the approach. The obtained results are promising for the use of this approach in a real network of smart cameras.

>Magister Scientiae - MSc / The SASL project is in the process of developing a machine translation system that can translate fully-fledged phrases between SASL and English in real-time. To-date, several systems have been developed by the project focusing on facial expression, hand shape, hand motion, hand orientation and hand location recognition and estimation. Achmed developed a highly accurate upper body pose recognition and estimation system. The system is capable of recognizing and estimating the location of the arms from a twodimensional video captured from a monocular view at an accuracy of 88%. The system operates at well below real-time speeds. This research aims to investigate the use of optimizations and parallel processing techniques using the CUDA framework on Achmed’s algorithm to achieve real-time upper body pose recognition and estimation. A detailed analysis of Achmed’s algorithm identified potential improvements to the algorithm. Are- implementation of Achmed’s algorithm on the CUDA framework, coupled with these improvements culminated in an enhanced upper body pose recognition and estimation system that operates in real-time with an increased accuracy.

Magister Scientiae - MSc / Research has shown that five fundamental parameters are required to recognize any sign language gesture: hand shape, hand motion, hand location, hand orientation and facial expressions. The South African Sign Language (SASL) research group at the University of the Western Cape (UWC) has created several systems to recognize sign language gestures using single parameters. These systems are, however, limited to a vocabulary size of 20 – 23 signs, beyond which the recognition accuracy is expected to decrease. The first aim of this research is to investigate the use of two parameters – hand motion and hand shape – to recognise a larger vocabulary of SASL gestures at a high accuracy. Also, the majority of related work in the field of sign language gesture recognition using these two parameters makes use of Hidden Markov Models (HMMs) to classify gestures. Hidden Markov Support Vector Machines (HM-SVMs) are a relatively new technique that make use of Support Vector Machines (SVMs) to simulate the functions of HMMs. Research indicates that HM-SVMs may perform better than HMMs in some applications. To our knowledge, they have not been applied to the field of sign language gesture recognition. This research compares the use of these two techniques in the context of SASL gesture recognition. The results indicate that, using two parameters results in a 15% increase in accuracy over the use of a single parameter. Also, it is shown that HM-SVMs are a more accurate technique than HMMs, generally performing better or at least as good as HMMs.

People counting system involves the process of counting and estimating the number of people in a scene. The counting system has a number of useful applications, ranging from pedestrian traffic surveillance and monitoring the number of people that enters and leaves shopping malls to commercial buildings, vehicles and a number of other security-related applications. Over the years, significant progress has been made. However, people counting systems still have not overcome a number of challenges such as occlusions, human pose and direction, multiple people detection, varying lighting and weather conditions. The aim of this research is to present an optimal solution that is invariant to the challenges. That is, the outcome of the results will not be affected by the challenges. Also, the solution will handle the trade-off between the counting accuracy and the time it takes to implement the counting process. As a result, a new background subtraction method known as single pixel method is proposed. This is where useful features are collected from each scene using frame difference method. Then, these features are reduced into single pixels. The single pixels are then used to estimate the total number of people in the scene. Furthermore, a virtual-line direction-estimation method is presented where the directions in which the people are heading are estimated prior to counting. AFRIKAANS : Mense-telstelsels behels die proses van die tel en die beraming van die aantal mense op ’n toneel. Die telstelsel het ’n aantal nuttige toepassings wat wissel van voetgangerverkeer toesig en die monitering van die aantal mense wat binnekom en verlaat tot winkelsentrums, kommersiële geboue, voertuie, en ’n aantal ander sekuriteit-verwante programme. Oor die jare is beduidende vordering gemaak. Daar is egter ’n aantal uitdagings wat mense-telstelsels nog nie oorkom het nie, soos afsluiting, menslike inhou en rigting, die opsporing van veelvoudige mense, wisselende beligting en weerstoestande. Die doel van hierdie navorsing is om ’n optimale oplossing aan te bied, wat invariant is teen die uitdagings. Met ander woorde, die uitdagings sal nie die resultate affekteer nie. Die oplossing sal ook die uitruil tussen die tel akkuraatheid en die implementeringstyd van die telproses hanteer. As gevolg hiervan, is ’n nuwe agtergrondaftrekkingsmetode, wat bekend staan as ’n enkele beeldelement metode, voorgestel. Dit is waar die nuttige funksies van elke toneel, met behulp van die raamverskilmetode ingesamel word. Dan word hierdie eienskappe in enkele beeldelemente verminder. Die enkele beeldelemente word dan gebruik om die totale aantal mense in die toneel te skat. Verder is daar van ’n virtuele-lyn rigting-skatting metode gebruik gemaak wat die rigtings waarin die mense beweeg vooraf beraam. / Dissertation (MEng)--University of Pretoria, 2013. / Electrical, Electronic and Computer Engineering / unrestricted

The thesis deals with the processing and detection of specific objects in the image on the Android mobile platform. The main objective of this work was to design and then implement a mobile application for Android operating system, which allows control of pipetting processes based on images from a mobile device camera. The OpenCV library is used in the application for image processing. The resulting application should serve primarily in laboratories as a tool for complete analysis of the pipetting process. The work is divided into two main chapters, which further consist of sections and smaller subsections. The first chapter is devoted to the theoretical analysis of this work. Here is also describes used technology, Android operating system, OpenCV library and important parts of image processing. The second chapter deals with the proposal and subsequent practical solution of this work. There is a proposal and the following procedure for solving this work, important techniques, methods of processing and analysis of the camera image. The conclusion of the thesis is an evaluation of the results of the solution of this work.

The aim of the master thesis is to derive and implement image porcessing methods for people detection and tracking in images or videos. The overall solution was chosen as a combination of modern approaches and methods which were recently presented. The proposed algorithm is able to create trajectory of the person moving in indoor building spaces even under influence of full or partial occlusion for a short period of time. The scene of interest is surveyed by a static camera having direct view on targets. Selected methods are implemented in C# programming language based on OpenCV library. Graphical user interface was created to show the final output of algorithm.