This dissertation work is describing the usage of the software tool Computer Vision System Toolbox to create applications in computer vision. At the beginning of the work is performed background research of image scanning and its representation by using colour models. It is followed by a description of epipolar geometry and lastly is stated a description of the Computer Vision System Toolbox. In the next section of the work we deal with setting of used Basler cameras and processing of the scanned image. The following is a description how to create applications for object detection and after this description, we get to know applications for creation of depth maps area.

The representation and manipulation of visual content in a computer vision system requires a suitable abstraction of raw visual content such as pixels in an image. In this thesis, we study region-based feature representations and in particular, hierarchical segmentations because they do make no assumptions about region granularity. Hierarchical segmentations create a large feature space that increases the cost of subsequent processing in computer vision systems. We introduce a segment filter to reduce the feature space of hierarchical segmentations by identifying unique regions in the images. The filter uses appearance-based properties of the regions and the structure of the segmentation for the selection of a small set of descriptive regions. The filter works in two phases: selection with a criteria based on relative region size and a sorting based on a variational criteria. The filter is applicable to any hierarchical segmentation algorithm, in particular to bottom-up and region growing approaches. We evaluate the filter's performance against an extensive set of ground-truth regions from a dataset containing image sequences with scenes of different complexity. We demonstrate a novel region-to-region image matching approach as a possible application of our segment filter. A reduced segmentation tree is reconstructed based on the set of regions provided by the filtering. The reduction of the feature space by the segment filter simplifies our region-to-region matching approach. The correspondences between regions from two different images is established by a similarity measure. We use a modified mutual information measurement to compute the similarity of regions. The identified region correspondences are refined using the reduced segmentation tree. Our region-to-region matching approach is evaluated with an extensive set of ground-truth correspondences. This evaluation shows the large potential of both, our filtering and our matching approach.

The representation and manipulation of visual content in a computer vision system requires a suitable abstraction of raw visual content such as pixels in an image. In this thesis, we study region-based feature representations and in particular, hierarchical segmentations because they do make no assumptions about region granularity. Hierarchical segmentations create a large feature space that increases the cost of subsequent processing in computer vision systems. We introduce a segment filter to reduce the feature space of hierarchical segmentations by identifying unique regions in the images. The filter uses appearance-based properties of the regions and the structure of the segmentation for the selection of a small set of descriptive regions. The filter works in two phases: selection with a criteria based on relative region size and a sorting based on a variational criteria. The filter is applicable to any hierarchical segmentation algorithm, in particular to bottom-up and region growing approaches. We evaluate the filter's performance against an extensive set of ground-truth regions from a dataset containing image sequences with scenes of different complexity. We demonstrate a novel region-to-region image matching approach as a possible application of our segment filter. A reduced segmentation tree is reconstructed based on the set of regions provided by the filtering. The reduction of the feature space by the segment filter simplifies our region-to-region matching approach. The correspondences between regions from two different images is established by a similarity measure. We use a modified mutual information measurement to compute the similarity of regions. The identified region correspondences are refined using the reduced segmentation tree. Our region-to-region matching approach is evaluated with an extensive set of ground-truth correspondences. This evaluation shows the large potential of both, our filtering and our matching approach.

The representation and manipulation of visual content in a computer vision system requires a suitable abstraction of raw visual content such as pixels in an image. In this thesis, we study region-based feature representations and in particular, hierarchical segmentations because they do make no assumptions about region granularity. Hierarchical segmentations create a large feature space that increases the cost of subsequent processing in computer vision systems. We introduce a segment filter to reduce the feature space of hierarchical segmentations by identifying unique regions in the images. The filter uses appearance-based properties of the regions and the structure of the segmentation for the selection of a small set of descriptive regions. The filter works in two phases: selection with a criteria based on relative region size and a sorting based on a variational criteria. The filter is applicable to any hierarchical segmentation algorithm, in particular to bottom-up and region growing approaches. We evaluate the filter's performance against an extensive set of ground-truth regions from a dataset containing image sequences with scenes of different complexity. We demonstrate a novel region-to-region image matching approach as a possible application of our segment filter. A reduced segmentation tree is reconstructed based on the set of regions provided by the filtering. The reduction of the feature space by the segment filter simplifies our region-to-region matching approach. The correspondences between regions from two different images is established by a similarity measure. We use a modified mutual information measurement to compute the similarity of regions. The identified region correspondences are refined using the reduced segmentation tree. Our region-to-region matching approach is evaluated with an extensive set of ground-truth correspondences. This evaluation shows the large potential of both, our filtering and our matching approach.

The representation and manipulation of visual content in a computer vision system requires a suitable abstraction of raw visual content such as pixels in an image. In this thesis, we study region-based feature representations and in particular, hierarchical segmentations because they do make no assumptions about region granularity. Hierarchical segmentations create a large feature space that increases the cost of subsequent processing in computer vision systems. We introduce a segment filter to reduce the feature space of hierarchical segmentations by identifying unique regions in the images. The filter uses appearance-based properties of the regions and the structure of the segmentation for the selection of a small set of descriptive regions. The filter works in two phases: selection with a criteria based on relative region size and a sorting based on a variational criteria. The filter is applicable to any hierarchical segmentation algorithm, in particular to bottom-up and region growing approaches. We evaluate the filter's performance against an extensive set of ground-truth regions from a dataset containing image sequences with scenes of different complexity. We demonstrate a novel region-to-region image matching approach as a possible application of our segment filter. A reduced segmentation tree is reconstructed based on the set of regions provided by the filtering. The reduction of the feature space by the segment filter simplifies our region-to-region matching approach. The correspondences between regions from two different images is established by a similarity measure. We use a modified mutual information measurement to compute the similarity of regions. The identified region correspondences are refined using the reduced segmentation tree. Our region-to-region matching approach is evaluated with an extensive set of ground-truth correspondences. This evaluation shows the large potential of both, our filtering and our matching approach.

This dissertation addresses the problem of human detection and tracking in surveillance videos. Even though this is a well-explored topic, many challenges remain when confronted with data from real world situations. These challenges include appearance variation, illumination changes, camera motion, cluttered scenes and occlusion. In this dissertation several novel methods for improving on the current state of human detection and tracking based on learning scene-specific information in video feeds are proposed. Firstly, we propose a novel method for human detection which employs unsupervised learning and superpixel segmentation. The performance of generic human detectors is usually degraded in unconstrained video environments due to varying lighting conditions, backgrounds and camera viewpoints. To handle this problem, we employ an unsupervised learning framework that improves the detection performance of a generic detector when it is applied to a particular video. In our approach, a generic DPM human detector is employed to collect initial detection examples. These examples are segmented into superpixels and then represented using Bag-of-Words (BoW) framework. The superpixel-based BoW feature encodes useful color features of the scene, which provides additional information. Finally a new scene-specific classifier is trained using the BoW features extracted from the new examples. Compared to previous work, our method learns scene-specific information through superpixel-based features, hence it can avoid many false detections typically obtained by a generic detector. We are able to demonstrate a significant improvement in the performance of the state-of-the-art detector. Given robust human detection, we propose a robust multiple-human tracking framework using a part-based model. Human detection using part models has become quite popular, yet its extension in tracking has not been fully explored. Single camera-based multiple-person tracking is often hindered by difficulties such as occlusion and changes in appearance. We address such problems by developing an online-learning tracking-by-detection method. Our approach learns part-based person-specific Support Vector Machine (SVM) classifiers which capture articulations of moving human bodies with dynamically changing backgrounds. With the part-based model, our approach is able to handle partial occlusions in both the detection and the tracking stages. In the detection stage, we select the subset of parts which maximizes the probability of detection. This leads to a significant improvement in detection performance in cluttered scenes. In the tracking stage, we dynamically handle occlusions by distributing the score of the learned person classifier among its corresponding parts, which allows us to detect and predict partial occlusions and prevent the performance of the classifiers from being degraded. Extensive experiments using the proposed method on several challenging sequences demonstrate state-of-the-art performance in multiple-people tracking. Next, in order to obtain precise boundaries of humans, we propose a novel method for multiple human segmentation in videos by incorporating human detection and part-based detection potential into a multi-frame optimization framework. In the first stage, after obtaining the superpixel segmentation for each detection window, we separate superpixels corresponding to a human and background by minimizing an energy function using Conditional Random Field (CRF). We use the part detection potentials from the DPM detector, which provides useful information for human shape. In the second stage, the spatio-temporal constraints of the video is leveraged to build a tracklet-based Gaussian Mixture Model for each person, and the boundaries are smoothed by multi-frame graph optimization. Compared to previous work, our method could automatically segment multiple people in videos with accurate boundaries, and it is robust to camera motion. Experimental results show that our method achieves better segmentation performance than previous methods in terms of segmentation accuracy on several challenging video sequences. Most of the work in Computer Vision deals with point solution; a specific algorithm for a specific problem. However, putting different algorithms into one real world integrated system is a big challenge. Finally, we introduce an efficient tracking system, NONA, for high-definition surveillance video. We implement the system using a multi-threaded architecture (Intel Threading Building Blocks (TBB)), which executes video ingestion, tracking, and video output in parallel. To improve tracking accuracy without sacrificing efficiency, we employ several useful techniques. Adaptive Template Scaling is used to handle the scale change due to objects moving towards a camera. Incremental Searching and Local Frame Differencing are used to resolve challenging issues such as scale change, occlusion and cluttered backgrounds. We tested our tracking system on a high-definition video dataset and achieved acceptable tracking accuracy while maintaining real-time performance.

This master‘s thesis describes the design and implementation of the system for detection and recognition of speed limit road signs. It focuses on the recognition of the red circular speed limit sign from the image data using the computer vision methods. Several methods were programmed and tested as a part of this thesis. In the final solution, the segmentation based on YCbCr color model is used. Detection of the circular sign and final classification is performed by template matching method. Algorithm for the tracking of the detected signs between frames of the video is used for better performance in real-time recognition. Application is developed using MATLAB and Simulink. The result is a simple driver assistance system prototype, which can be implemented in any computer with camera. The correct function of the algorithm was confirmed during a testing in a traffic.

A reação álcali-agregado - RAA é uma patologia de ação lenta que tem sido observada em construções de concreto capaz de comprometer suas estruturas. Sabe-se que a reação álcali-agregado é um fenômeno bastante complexo em virtude da grande variedade de rochas na natureza que são empregadas como agregados no preparo do concreto, podendo cada mineral utilizado afetar de forma distinta a reação ocorrida. Em função dos tipos de estrutura, das suas condições de exposição e dos materiais empregados, a RAA não se comporta sempre da mesma forma, em virtude disto a pesquisa constante neste tema é necessária para o meio técnico e a sociedade. Pesquisas laboratoriais, empíricas e experimentais tem sido rotina em muitos dos estudos da RAA dada ainda à carência de certas definições mais precisas a respeito dos métodos de ensaio, mas também em função da necessidade do melhor conhecimento dos materiais de uso em concretos como os agregados, cimentos, adições, aditivos entre outros e do comportamento da estrutura. Embora técnicas de prevenção possam reduzir significativamente a incidência da RAA, muitas estruturas foram construídas antes que tais medidas fossem conhecidas, havendo no Brasil vários casos de estruturas afetadas, sendo custosos os reparos dessas estruturas. Em estudos recentes sobre o tamanho das partículas de álcali-agregado e sua distribuição foi concluído que o tamanho do agregado está relacionado com o potencial danoso da RAA. Existem ainda indícios de que o tamanho e a distribuição dos poros do concreto também sejam capazes de influenciar o potencial reativo do concreto. Neste trabalho desenvolvemos um Sistema de Visão Artificial (SVA) que, com o uso de técnicas de Processamento de Imagens, é capaz de identificar em imagens de concreto, agregado e poros que atendam em sua forma, às especificações do usuário, possibilitando o cálculo da porosidade e produzindo imagens segmentadas à partir das quais será possível extrair dados relativos à geometria desses elementos. Serão feitas duas abordagens para a obtenção das imagens, uma por Escâner Comercial, que possui vantagens relacionadas à facilidade de aquisição do equipamento, e outra por micro tomógrafo. Uma vez obtidas informações sobre as amostras de concreto, estas podem ser utilizadas para pesquisar a RAA, comparar estruturas de risco com estruturas antigas de forma a melhorar a previsão de risco de ocorrência, bem como serem aplicadas a outras no estudo de outras patologias do concreto menos comuns no nosso país, como o efeito gelo/degelo. / The alkali-aggregate reaction - RAA is a condition of slow action that has been observed in concrete constructions that could affect their structures. It is known that the alkali-aggregate reaction is a very complex phenomenon because of the great variety of rocks in nature that are used as aggregates for concrete, and each mineral used differently affects the reaction occurred. Depending on the type of structure, its exposure conditions and the materials used, this phenomenon does not always behaves the same way, because of this, constant research in this area is needed for the technical means and the society. Laboratory, empirical and experimental research has been routine in many of the RAA studies still given the lack of certain more precise definitions concerning the testing methods, but also because of the need for better understanding of the use of materials in concrete as aggregate, cement, additions, additives etc. and structure behavior. Prevention techniques could significantly reduce the incidence of RAA. Still, many structures were built before such measures were known, several cases of affected structures were discovered in Brazil, all with large spending on repairs of the affected structures. In recent studies on the particle size of the alkaliaggregate and its distribution was concluded that the aggregate size is related to the damaging potential of the RAA. There are also indications that the size and distribution of concrete pores are also capable of influencing the reactive potential of the concrete. In the present work we developed an Artificial Vision System ( VAS ) that uses image processing techniques to identify aggregate and pores in hardened concrete images, enabling the calculation of porosity and producing segmented images that can be used to investigate data about the geometry of these elements. Were made two approaches for obtaining the images, one by Scanner Commercial, which has related advantages will ease the acquisition of equipment, and other micro CT scanner. Once obtained information on the concrete samples, these can be used to search the AAR compared risk structures with old structures so as to enhance the occurrence of risk prediction, as well as be applied to other concrete in the study of other pathologies less common in our country, as ice effect / thaw.

A reação álcali-agregado - RAA é uma patologia de ação lenta que tem sido observada em construções de concreto capaz de comprometer suas estruturas. Sabe-se que a reação álcali-agregado é um fenômeno bastante complexo em virtude da grande variedade de rochas na natureza que são empregadas como agregados no preparo do concreto, podendo cada mineral utilizado afetar de forma distinta a reação ocorrida. Em função dos tipos de estrutura, das suas condições de exposição e dos materiais empregados, a RAA não se comporta sempre da mesma forma, em virtude disto a pesquisa constante neste tema é necessária para o meio técnico e a sociedade. Pesquisas laboratoriais, empíricas e experimentais tem sido rotina em muitos dos estudos da RAA dada ainda à carência de certas definições mais precisas a respeito dos métodos de ensaio, mas também em função da necessidade do melhor conhecimento dos materiais de uso em concretos como os agregados, cimentos, adições, aditivos entre outros e do comportamento da estrutura. Embora técnicas de prevenção possam reduzir significativamente a incidência da RAA, muitas estruturas foram construídas antes que tais medidas fossem conhecidas, havendo no Brasil vários casos de estruturas afetadas, sendo custosos os reparos dessas estruturas. Em estudos recentes sobre o tamanho das partículas de álcali-agregado e sua distribuição foi concluído que o tamanho do agregado está relacionado com o potencial danoso da RAA. Existem ainda indícios de que o tamanho e a distribuição dos poros do concreto também sejam capazes de influenciar o potencial reativo do concreto. Neste trabalho desenvolvemos um Sistema de Visão Artificial (SVA) que, com o uso de técnicas de Processamento de Imagens, é capaz de identificar em imagens de concreto, agregado e poros que atendam em sua forma, às especificações do usuário, possibilitando o cálculo da porosidade e produzindo imagens segmentadas à partir das quais será possível extrair dados relativos à geometria desses elementos. Serão feitas duas abordagens para a obtenção das imagens, uma por Escâner Comercial, que possui vantagens relacionadas à facilidade de aquisição do equipamento, e outra por micro tomógrafo. Uma vez obtidas informações sobre as amostras de concreto, estas podem ser utilizadas para pesquisar a RAA, comparar estruturas de risco com estruturas antigas de forma a melhorar a previsão de risco de ocorrência, bem como serem aplicadas a outras no estudo de outras patologias do concreto menos comuns no nosso país, como o efeito gelo/degelo. / The alkali-aggregate reaction - RAA is a condition of slow action that has been observed in concrete constructions that could affect their structures. It is known that the alkali-aggregate reaction is a very complex phenomenon because of the great variety of rocks in nature that are used as aggregates for concrete, and each mineral used differently affects the reaction occurred. Depending on the type of structure, its exposure conditions and the materials used, this phenomenon does not always behaves the same way, because of this, constant research in this area is needed for the technical means and the society. Laboratory, empirical and experimental research has been routine in many of the RAA studies still given the lack of certain more precise definitions concerning the testing methods, but also because of the need for better understanding of the use of materials in concrete as aggregate, cement, additions, additives etc. and structure behavior. Prevention techniques could significantly reduce the incidence of RAA. Still, many structures were built before such measures were known, several cases of affected structures were discovered in Brazil, all with large spending on repairs of the affected structures. In recent studies on the particle size of the alkaliaggregate and its distribution was concluded that the aggregate size is related to the damaging potential of the RAA. There are also indications that the size and distribution of concrete pores are also capable of influencing the reactive potential of the concrete. In the present work we developed an Artificial Vision System ( VAS ) that uses image processing techniques to identify aggregate and pores in hardened concrete images, enabling the calculation of porosity and producing segmented images that can be used to investigate data about the geometry of these elements. Were made two approaches for obtaining the images, one by Scanner Commercial, which has related advantages will ease the acquisition of equipment, and other micro CT scanner. Once obtained information on the concrete samples, these can be used to search the AAR compared risk structures with old structures so as to enhance the occurrence of risk prediction, as well as be applied to other concrete in the study of other pathologies less common in our country, as ice effect / thaw.

Driving is one of the most common and preferred forms of transport used in our actual society. However, according to studies, it is also one of the most dangerous. One solution to increase safety on the road is applying technology to automate and prevent avoidable human errors. Nevertheless, despite the efforts to obtain reliable systems, we have yet to find a reliable and safe enough solution for solving autonomous driving. One of the reasons is that many drives are done in conditions far from the ideal, with variable lighting conditions and fast-paced, unpredictable environments. This project develops and evaluates an algorithm that takes the input of dynamic vision sensors (DVS) and runs on neuromorphic spiking neural networks (SNN) to obtain a robust road lane tracking system. We present quantitative and qualitative metrics that evaluate the performance of lane recognition in low light conditions against conventional algorithms. This project is motivated by the main advantages of neuromorphic vision sensors: recognizing a high dynamic range and allowing a high-speed image capture. Another improvement of this system is the computational speed and power efficiency that characterize neuromorphic hardware based on spiking neural networks. The results obtained show a similar accuracy of this new algorithm compared to previous implementations on conventional hardware platforms. Most importantly, it accomplishes the proposed task with lower latency and computing power requirements than previous algorithms. / Att köra bil är ett av de vanligaste och mest populära transportsätten i vårt samhälle. Enligt forskningen är det också ett av de farligaste. En lösning för att öka säkerheten på vägarna är att med teknikens hjälp automatisera bilkörningen och på så sätt förebygga misstag som beror på den mänskliga faktorn. Trots ansträngningarna för att få fram tillförlitliga system har man dock ännu inte hittat en tillräckligt tillförlitlig och säker lösning för självkörande bilar. En av orsakerna till det är att många körningar sker under förhållanden som är långt ifrån idealiska, med varierande ljusförhållanden och oförutsägbara miljöer i höga hastigheter. I det här projektet utvecklar och utvärderar vi en algoritm som tar emot indata från dynamiska synsensorer (Dynamic Vision Sensors, DVS) och kör datan på neuromorfiska pulserande neuronnät (Spiking Neural Networks, SNN) för att skapa ett robust system för att läsa av vägbanan. Vi presenterar en kvantitativ och kvalitativ utvärdering av hur väl systemet läser av körbanans linjer i svagt ljus, och jämför därefter resultaten med dem för tidigare algoritmer. Detta projekt motiveras av de viktigaste fördelarna med neuromorfiska synsensorer: brett dynamiskt omfång och hög bildtagningshastighet. En annan fördel hos detta system är den korta beräkningstiden och den energieffektivitet som kännetecknar neuromorfisk hårdvara baserad på pulserande neuronnät. De resultat som erhållits visar att den nya algoritmen har en liknande noggrannhet som tidigare algoritmer på traditionella hårdvaruplattformar. I jämförelse med den traditionella tekniken, utför algoritmen i den föreliggande studien sin uppgift med kortare latenstid och lägre krav på processorkraft. / La conducción es una de las formas de transporte más comunes y preferidas en la actualidad. Sin embargo, diferentes estudios muestran que también es una de las más peligrosas. Una solución para aumentar la seguridad en la carretera es aplicar la tecnología para automatizar y prevenir los evitables errores humanos. No obstante, a pesar de los esfuerzos por conseguir sistemas fiables, todavía no hemos encontrado una solución suficientemente fiable y segura para resolver este reto. Una de las razones es el entorno de la conducción, en situaciones que distan mucho de las ideales, con condiciones de iluminación variables y entornos rápidos e imprevisibles. Este proyecto desarrolla y evalúa un algoritmo que toma la entrada de sensores de visión dinámicos (DVS) y ejecuta su computación en redes neuronales neuromórficas (SNN) para obtener un sistema robusto de seguimiento de carriles en carretera. Presentamos métricas cuantitativas y cualitativas que evalúan el rendimiento del reconocimiento de carriles en condiciones de poca luz, frente a algoritmos convencionales. Este proyecto está motivado por la validación de las ventajas de los sensores de visión neuromórficos: el reconocimiento de un alto rango dinámico y la captura de imágenes de alta velocidad. Otra de las mejoras que se espera de este sistema es la velocidad de procesamiento y la eficiencia energética que caracterizan al hardware neuromórfico basado en redes neuronales de impulsos. Los resultados obtenidos muestran una precisión similar entre el nuevo algoritmo en comparación con implementaciones anteriores en plataformas convencionales. Y lo que es más importante, realiza la tarea propuesta con menor latencia y requisitos de potencia de cálculo.