Siga este enlace para ver otros tipos de publicaciones sobre el tema: Natural images.
Tesis sobre el tema "Natural images"
Crea una cita precisa en los estilos APA, MLA, Chicago, Harvard y otros
Consulte los 50 mejores tesis para su investigación sobre el tema "Natural images".
Junto a cada fuente en la lista de referencias hay un botón "Agregar a la bibliografía". Pulsa este botón, y generaremos automáticamente la referencia bibliográfica para la obra elegida en el estilo de cita que necesites: APA, MLA, Harvard, Vancouver, Chicago, etc.
También puede descargar el texto completo de la publicación académica en formato pdf y leer en línea su resumen siempre que esté disponible en los metadatos.
Explore tesis sobre una amplia variedad de disciplinas y organice su bibliografía correctamente.
1
Chen, Ting-Li. "On the statistics of natural images /". View online version; access limited to Brown University users, 2005. http://wwwlib.umi.com/dissertations/fullcit/3174586.
Texto completo
2
Miflah, Hussain Ismail Ahamed. "Higher-level representations of natural images". Thesis, Queen Mary, University of London, 2018. http://qmro.qmul.ac.uk/xmlui/handle/123456789/39759.
Texto completoResumen
The traditional view of vision is that neurons in early cortical areas process information about simple features (e.g. orientation and spatial frequency) in small, spatially localised regions of visual space (the neuron's receptive field). This piecemeal information is then fed-forward into later stages of the visual system where it gets combined to form coherent and meaningful global (higher-level) representations. The overall aim of this thesis is to examine and quantify this higher level processing; how we encode global features in natural images and to understand the extent to which our perception of these global representations is determined by the local features within images. Using the tilt after-effect as a tool, the first chapter examined the processing of a low level, local feature and found that the orientation of a sinusoidal grating could be encoded in both a retinally and spatially non-specific manner. Chapter 2 then examined these tilt aftereffects to the global orientation of the image (i.e., uprightness). We found that image uprightness was also encoded in a retinally / spatially non-specific manner, but that this global property could be processed largely independently of its local orientation content. Chapter 3 investigated if our increased sensitivity to cardinal (vertical and horizontal) structures compared to inter-cardinal (45° and 135° clockwise of vertical) structures, influenced classification of unambiguous natural images. Participants required relatively less contrast to classify images when they retained near-cardinal as compared to near-inter-cardinal structures. Finally, in chapter 4, we examined category classification when images were ambiguous. Observers were biased to classify ambiguous images, created by combining structures from two distinct image categories, as carpentered (e.g., a house). This could not be explained by differences in sensitivity to local structures and is most likely the result of our long-term exposure to city views. Overall, these results show that higher-level representations are not fully dependent on the lower level features within an image. Furthermore, our knowledge about the environment influences the extent to which we use local features to rapidly identify an image.
3
Tavakoli, Fatemeh. "On Visual Attention in Natural Images". Thesis, Linnéuniversitetet, Institutionen för fysik och elektroteknik (IFE), 2015. http://urn.kb.se/resolve?urn=urn:nbn:se:lnu:diva-48256.
Texto completoResumen
By visual attention process biological and machine vision systems are able to select the most relevant regions from a scene. The relevancy process is achieved either by top-down factors, driven by task, or bottom-up factors, the visual saliency, which distinguish a scene region that are different from its surrounding. During the past 20 years numerous research efforts have aimed to model bottom-up visual saliency with many successful applications in computer vision and robotics.In this thesis we have performed a comparison between a state-of-the-art saliency model and subjective test (human eye tracking) using different evaluation methods over three generated dataset of synthetic patterns and natural images. Our results showed that the objective model is partially valid and highly center-biased.By using empirical data obtained from subjective experiments we propose a special function, the Probability of Characteristic Radially Dependency Function, to model the lateral distribution of visual attention process.
4
Kim, Kyu-Heon. "Segmentation of natural texture images using a robust stochastic image model". Thesis, University of Newcastle Upon Tyne, 1996. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.307927.
Texto completo
5
Azzabou, Noura. "Variable Bandwidth Image Models for Texture-Preserving Enhancement of Natural Images". Paris Est, 2008. http://pastel.paristech.org/4041/01/ThesisNouraAzzabou.pdf.
Texto completoResumen
Cette thèse s’intéresse aux problèmes de restauration d’images et de préservation de textures. Cette tâche nécessite un modèle image qui permet de caractériser le signal qu’on doit obtenir. Un tel model s’appuie sur la définition de l’interaction entre les pixels et qui est caractérisé par deux aspects : (1) la similarité photométrique entre les pixels (2) la distance spatiale entre les pixels qui peut être comparée à une grandeur d’échelle. La première partie de la thèse introduit un nouveau modèle non paramétrique d’image. Ce modèle permet d’obtenir une description adaptative de l’image en utilisant des noyaux de taille variable obtenue à partir d’une étape de classification effectuée au préalable. La deuxième partie introduit une autre approche pour décrire la dépendance entre pixels d’un point de vue géométrique. Ceci est effectué à l’aide d’un modèle statistique de la co-occurrence entre les observations de point de vue géométrique. La dernière partie est une nouvelle technique de sélection automatique (pour chaque pixel) de la taille des noyaux utilisé au cours du filtrage. Cette thèse est conclue avec l’application de cette dernière approche dans différents contextes de filtrage ce qui montre sa flexibilité vis-à-vis des contraintes liées aux divers problèmes traitésThis thesis is devoted to image enhancement and texture preservation issues. This task involves an image model that describes the characteristics of the recovered signal. Such a model is based on the definition of the pixels interaction that is often characterized by two aspects (i) the photometric similarity between pixels (ii) the spatial distance between them that can be compared to a given scale. The first part of the thesis, introduces novel non-parametric image models towards more appropriate and adaptive image description using variable bandwidth approximations driven from a soft classification in the image. The second part introduces alternative means to model observations dependencies from geometric point of view. This is done through statistical modeling of co-occurrence between observations and the use of multiple hypotheses testing and particle filters. The last part is devoted to novel adaptive means for spatial bandwidth selection and more efficient tools to capture photometric relationships between observations. The thesis concludes with providing other application fields of the last technique towards proving its flexibility toward various problem requirements
6
Viklund, Alexander y Emma Nimstad. "Character Recognition in Natural Images Utilising TensorFlow". Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-208385.
Texto completoResumen
Convolutional Neural Networks (CNNs) are commonly used for character recognition. They achieve the lowest error rates for popular datasets such as SVHN and MNIST. Usage of CNN is lacking in research about character classification in natural images regarding the whole English alphabet. This thesis conducts an experiment where TensorFlow is used to construct a CNN that is trained and tested on the Chars74K dataset, with 15 images per class for training and 15 images per class for testing. This is done with the aim of achieving a higher accuracy than the non-CNN approach by de Campos et al. [1], that achieved 55.26%. The thesis explores data augmentation techniques for expanding the small training set and evaluates the result of applying rotation, stretching, translation and noise-adding. The result of this is that all of these methods apart from adding noise gives a positive effect on the accuracy of the network. Furthermore, the experiment shows that with a three layered convolutional neural network it is possible to create a character classifier that is as good as de Campos et al.'s. It is believed that even better results can be achieved if more experiments would be conducted on the parameters of the network and the augmentation.Det är vanligt att använda konvolutionära artificiella neuronnät (CNN) för bildigenkänning, då de ger de minsta felmarginalerna på kända datamängder som SVHN och MNIST. Dock saknas det forskning om användning av CNN för klassificering av bokstäver i naturliga bilder när det gäller hela det engelska alfabetet. Detta arbete beskriver ett experiment där TensorFlow används för att bygga ett CNN som tränas och testas med bilder från Chars74K. 15 bilder per klass används för träning och 15 per klass för testning. Målet med detta är att uppnå högre noggrannhet än 55.26%, vilket är vad de campos et al. [1] uppnådde med en metod utan artificiella neuronnät. I rapporten utforskas olika tekniker för att artificiellt utvidga den lilla datamängden, och resultatet av att applicera rotation, utdragning, translation och bruspåslag utvärderas. Resultatet av det är att alla dessa metoder utom bruspåslag ger en positiv effekt på nätverkets noggrannhet. Vidare visar experimentet att med ett CNN med tre lager går det att skapa en bokstavsklassificerare som är lika bra som de Campos et al.s klassificering. Om fler experiment skulle genomföras på nätverkets och utvidgningens parametrar är det troligt att ännu bättre resultat kan uppnås.
7
Granlund, Oskar y Kai Böhrnsen. "Improving character recognition by thresholding natural images". Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-208899.
Texto completoResumen
The current state of the art optical character recognition (OCR) algorithms are capable of extracting text from images in predefined conditions. OCR is extremely reliable for interpreting machine-written text with minimal distortions, but images taken in a natural scene are still challenging. In recent years the topic of improving recognition rates in natural images has gained interest because more powerful handheld devices are used. The main problem faced dealing with recognition in natural images are distortions like illuminations, font textures, and complex backgrounds. Different preprocessing approaches to separate text from its background have been researched lately. In our study, we assess the improvement reached by two of these preprocessing methods called k-means and Otsu by comparing their results from an OCR algorithm. The study showed that the preprocessing made some improvement on special occasions, but overall gained worse accuracy compared to the unaltered images.Dagens optisk teckeninläsnings (OCR) algoritmer är kapabla av att extrahera text från bilder inom fördefinierade förhållanden. De moderna metoderna har uppnått en hög träffsäkerhet för maskinskriven text med minimala förvrängningar, men bilder tagna i en naturlig scen är fortfarande svåra att hantera. De senaste åren har ett stort intresse för att förbättra tecken igenkännings algoritmerna uppstått, eftersom fler kraftfulla och handhållna enheter används. Det huvudsakliga problemet när det kommer till igenkänning i naturliga bilder är olika förvrängningar som infallande ljus, textens textur och komplicerade bakgrunder. Olika metoder för förbehandling och därmed separation av texten och dess bakgrund har studerats under den senaste tiden. I våran studie bedömer vi förbättringen som uppnås vid förbehandlingen med två metoder som kallas för k-means och Otsu genom att jämföra svaren från en OCR algoritm. Studien visar att Otsu och k-means kan förbättra träffsäkerheten i vissa förhållanden men generellt sett ger det ett sämre resultat än de oförändrade bilderna.
8
Johnson, Samuel Alan. "Articulated human pose estimation in natural images". Thesis, University of Leeds, 2012. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.598026.
Texto completoResumen
In this thesis the problem of estimating the 2-D articulated pose, or configuration of a person in unconstrained images such as consumer photographs is addressed. Contributions are split among three major chapters. In previous work the Pictorial Structure Model approach has proven particularly successful. and is appealing because of its moderate computational cost. However, the accuracy of resulting pose estimates has been limited by the use of simple representations of limb appearance. In this thesis strong discriminatively trained limb detectors combining gradient and colour segmentation cues are proposed. The approach improves significantly on the "iterative image parsing" method which was the state-of-the-art at the time, and shows significant promise for combination with other models of pose and appearance. In the second pan of this thesis higher fidelity models of pose and appearance are proposed. The aim is to tackle extremely challenging properties of the human pose estimation task arising from variation in pose, anatomy, clothing. and imaging conditions. Current methods use simple models of body part appearance and plausible configurations due to limitations of available training data and constraints on computational expense. It is shown that such models severely limit accuracy. A new annotated database of challenging consumer images is introduced, an order of magnitude larger than currently available datasets. This larger amount of data allows partitioning of the pose space and the learning of multiple, clustered Pictorial Structure Models. A relative improvement in accuracy of over 50% is achieved compared to the standard, single model approach. In the final part of this thesis the clustered Pictorial Structure Model framework is extended to handle much larger quantities of training data. Furthermore it is shown how to utilise Amazon Mechanical Turk and a latent annotation update scheme to achieve high quality annotations at low cost. A significant increase in pose estimation accuracy is presented, while the computational expense of the framework is improved by a factor of
9
Nasrallah, Alexandre James. "Statistics of gradient directions in natural images". Thesis, University College London (University of London), 2007. http://discovery.ucl.ac.uk/1444886/.
Texto completoResumen
Interest in finding statistical regularities in natural images has been growing since the advent of information theory and the advancement of the efficient coding hypothesis that the human visual system is optimised to encode natural visual stimuli. In this thesis, a statistical analysis of gradient directions in an ensemble of natural images is reported. Information-theoretic measures have been used to compute the amount of dependency which exists between triples of gradient directions at separate image locations. Control experiments are performed on other image classes: phase randomized natural images, whitened natural images, and Gaussian noise images. The main results show that for an ensemble of natural images the average amount of de pendency between two and three gradient directions is the same as for an ensemble of phase randomized natural images. This result does not extend to i) the amount dependency between gradient magnitudes, ii) gradient directions at high gradient magnitude locations, or iii) individual natural images. Furthermore, no significant synergetic dependencies are found between triples of gradient directions in an ensemble natural images a synergetic dependency is an increase in dependency between a pair of gradient directions given the interaction of a third gradient direction. Additional experiments are performed to establish both the generality and specificity of the main results by studying the gradient direction dependencies of ensembles of noise (random phases) images with varying power law power spectra. The results of the additional experiments indicate that, for ensembles of images with varying power law power spectra, the amount of dependency between two and three gradient directions is determined by the ensemble's mean power spectrum rather than the phase spectrum. A framework is also presented for future work and preliminary results are provided for the dependency between second order derivative measurements (shape index) for up to 9-point configurations.
10
Ma, Yufeng. "Going Deeper with Images and Natural Language". Diss., Virginia Tech, 2019. http://hdl.handle.net/10919/99993.
Texto completoResumen
One aim in the area of artificial intelligence (AI) is to develop a smart agent with high intelligence that is able to perceive and understand the complex visual environment around us. More ambitiously, it should be able to interact with us about its surroundings in natural languages. Thanks to the progress made in deep learning, we've seen huge breakthroughs towards this goal over the last few years. The developments have been extremely rapid in visual recognition, in which machines now can categorize images into multiple classes, and detect various objects within an image, with an ability that is competitive with or even surpasses that of humans. Meanwhile, we also have witnessed similar strides in natural language processing (NLP). It is quite often for us to see that now computers are able to almost perfectly do text classification, machine translation, etc. However, despite much inspiring progress, most of the achievements made are still within one domain, not handling inter-domain situations. The interaction between the visual and textual areas is still quite limited, although there has been progress in image captioning, visual question answering, etc.
In this dissertation, we design models and algorithms that enable us to build in-depth connections between images and natural languages, which help us to better understand their inner structures. In particular, first we study how to make machines generate image descriptions that are indistinguishable from ones expressed by humans, which as a result also achieved better quantitative evaluation performance. Second, we devise a novel algorithm for measuring review congruence, which takes an image and review text as input and quantifies the relevance of each sentence to the image. The whole model is trained without any supervised ground truth labels. Finally, we propose a brand new AI task called Image Aspect Mining, to detect visual aspects in images and identify aspect level rating within the review context.
On the theoretical side, this research contributes to multiple research areas in Computer Vision (CV), Natural Language Processing (NLP), interactions between CVandNLP, and Deep Learning. Regarding impact, these techniques will benefit related users such as the visually impaired, customers reading reviews, merchants, and AI researchers in general.Doctor of Philosophy
11
Sakellariou, Georgios. "Hybrid scene characterisation applied to natural images". Thesis, Imperial College London, 2012. http://hdl.handle.net/10044/1/9655.
Texto completoResumen
In this thesis, a combination of skeletonisation and graph matching techniques, coupled with a blend of supervised and unsupervised learning methodology is applied to the task of characterising and classifying natural shapes. A novel navigation-based skeletonisation algorithm is used to gather low level structural and morphological information about the shape. Subsequently, the data are converted into a series of attributed graphs, which characterise the image. Graphs of the same type can then be compared using an approximate graph matcher, which identifies a degree of similarity between them. Each degree of similarity corresponds to a data point in a conceptual space (as defined by Gärdenfors). The proposed method is applied to two distinct problems; the classification of leaf types, and the characterisation of river networks. The classification and characterisation systems are tested on a database of images of leaves and a collection of satellite images respectively. The novel navigation-based skeletonisation algorithm features several advantages; first, it allows the collection of topological and morphological information on the fly. This eliminates the need for any post-processing on the extracted skeletons. In addition, the adaptation of the algorithm to suit different applications is facilitated by the fact that any sort of morphological information can be included without alterations to the function of the algorithm. The conversion of the skeletons to attributed graphs is simplified by the existence of structural and morphological flags in the skeletal points. Last, concepts are created in the resulting conceptual space by means of a best-guess approach as well as a mechanism for accommodating external user input.
12
Lundqvist, Melvin y Agnes Forsberg. "A comparison of OCR methods on natural images in different image domains". Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-280286.
Texto completoResumen
Optical character recognition (OCR) is a blanket term for methods that convert printed or handwritten text into machine-encoded text. As the digital world keeps growing the amount of digital images with text increases, and the need for OCR methods that can handle more than plain text documents as well. There are OCR engines that can convert images of clean documents with an over 99% recognition rate. OCR for natural images is getting more and more attention, but because natural images can be far more diverse than plain text documents it also leads to complications. To combat these issues it needs to be clear in what areas the OCR methods of today struggle. This thesis aims to answer this by testing three popular, readily available, OCR methods on a dataset comprised only of natural images containing text. The results show that one of the methods, GOCR, can not handle natural images as its test results were very far from correct. For the other two methods, ABBYY FineReader and Tesseract, the results were better but also show that there still is a long way to go, especially when it comes to images with special font. However when the images are less complicated some of our methods performed above our expectations.Optical character recognition (OCR) är en samlingsterm för metoder som konverterar tryckt eller handskriven text till maskinkod. När den digitala världen växer så växer även antalet digitala bilder med text, och även behovet för OCR metoder som kan hantera mer än vanliga textdokument. Det finns idag OCR motorer som kan konvertera bilder av rena dokument till maskinkod med över 99% korrekthet. OCR för fotografier får mer och mer uppmärksamhet, men eftersom fotografier har mycket större mångfaldhet än rena textdokument leder detta också till problem. För att hantera detta krävs klarhet inom vilka områden som dagens OCR-metoder har problem. Denna uppsats ämnar svara på denna fråga genom att undersöka och testa tre populära, enkelt tillgängliga OCR metoder på ett dataset som endast innehåller fotografier av naturliga miljöer med text. Resultaten visade att en av metoderna, GOCR, inte kan hantera fotografier. GOCRs testresultat var långt från det korrekta. För de andra metoderna, ABBYY FineReader och Tesseract, var resultaten bättre men visade att det fortfarande finns mycket arbete att göra inom området, särskilt när det kommer till bilder med speciella typsnitt. När det däremot kommer till bilder som är mindre komplicerade blev vi förvånade över hur bra resultatet var för några av metoderna.
13
Christou, Christakis Georgiou. "Human vision and the physics of natural images". Thesis, University of Oxford, 1994. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.386936.
Texto completo
14
Mohammad, Nassir. "Sparse modelling of natural images and compressive sensing". Thesis, Cardiff University, 2011. http://orca.cf.ac.uk/55097/.
Texto completoResumen
This thesis concerns the study of the statistics of natural images and compressive sensing for two main objectives: 1) to extend our understanding of the regularities exhibited by natural images of the visual world we regularly view around us, and 2) to incorporate this knowledge into image processing applications. Previous work on image statistics has uncovered remarkable behavior of the dis tributions obtained from filtering natural images. Typically we observe high kurtosis, non-Gaussian distributions with sharp central cusps, which are called sparse in the literature. These results have become an accepted fact through empirical findings us ing zero mean filters on many different databases of natural scenes. The observations have played an important role in computational and biological applications, where re searchers have sought to understand visual processes through studying the statistical properties of the objects that are being observed. Interestingly, such results on sparse distributions also share elements with the emerging field of compressive sensing. This is a novel sampling protocol where one seeks to measure a signal in already com pressed format through randomised projections, while the recovery algorithm consists of searching for a constrained solution with the sparsest transformed coefficients. In view of prior art, we extend our knowledge of image statistics from the monochrome domain into the colour domain. We study sparse response distributions of filters constructed on colour channels and observe the regularity of the distributions across diverse datasets of natural images. Several solutions to image processing problems emerge from the incorporation of colour statistics as prior information. We give a Bayesian treatment to the problem of colorizing natural gray images, and formulate image compression schemes using elements of compressive sensing and sparsity. We also propose a denoising algorithm that utilises the sparse filter responses as a regular- isation function for the effective attenuation of Gaussian and impulse noise in images. The results emanating from this body of work illustrate how the statistics of natural images, when incorporated with Bayesian inference and sparse recovery, can have deep implications for image processing applications.
15
Hold-Geoffroy, Yannick. "Learning geometric and lighting priors from natural images". Doctoral thesis, Université Laval, 2018. http://hdl.handle.net/20.500.11794/31264.
Texto completoResumen
Comprendre les images est d’une importance cruciale pour une pléthore de tâches, de la composition numérique au ré-éclairage d’une image, en passant par la reconstruction 3D d’objets. Ces tâches permettent aux artistes visuels de réaliser des chef-d’oeuvres ou d’aider des opérateurs à prendre des décisions de façon sécuritaire en fonction de stimulis visuels. Pour beaucoup de ces tâches, les modèles physiques et géométriques que la communauté scientifique a développés donnent lieu à des problèmes mal posés possédant plusieurs solutions, dont généralement une seule est raisonnable. Pour résoudre ces indéterminations, le raisonnement sur le contexte visuel et sémantique d’une scène est habituellement relayé à un artiste ou un expert qui emploie son expérience pour réaliser son travail. Ceci est dû au fait qu’il est généralement nécessaire de raisonner sur la scène de façon globale afin d’obtenir des résultats plausibles et appréciables. Serait-il possible de modéliser l’expérience à partir de données visuelles et d’automatiser en partie ou en totalité ces tâches ? Le sujet de cette thèse est celui-ci : la modélisation d’a priori par apprentissage automatique profond pour permettre la résolution de problèmes typiquement mal posés. Plus spécifiquement, nous couvrirons trois axes de recherche, soient : 1) la reconstruction de surface par photométrie, 2) l’estimation d’illumination extérieure à partir d’une seule image et 3) l’estimation de calibration de caméra à partir d’une seule image avec un contenu générique. Ces trois sujets seront abordés avec une perspective axée sur les données. Chacun de ces axes comporte des analyses de performance approfondies et, malgré la réputation d’opacité des algorithmes d’apprentissage machine profonds, nous proposons des études sur les indices visuels captés par nos méthodes.Understanding images is needed for a plethora of tasks, from compositing to image relighting, including 3D object reconstruction. These tasks allow artists to realize masterpieces or help operators to safely make decisions based on visual stimuli. For many of these tasks, the physical and geometric models that the scientific community has developed give rise to ill-posed problems with several solutions, only one of which is generally reasonable. To resolve these indeterminations, the reasoning about the visual and semantic context of a scene is usually relayed to an artist or an expert who uses his experience to carry out his work. This is because humans are able to reason globally on the scene in order to obtain plausible and appreciable results. Would it be possible to model this experience from visual data and partly or totally automate tasks? This is the topic of this thesis: modeling priors using deep machine learning to solve typically ill-posed problems. More specifically, we will cover three research axes: 1) surface reconstruction using photometric cues, 2) outdoor illumination estimation from a single image and 3) camera calibration estimation from a single image with generic content. These three topics will be addressed from a data-driven perspective. Each of these axes includes in-depth performance analyses and, despite the reputation of opacity of deep machine learning algorithms, we offer studies on the visual cues captured by our methods.
16
Tadmor, Yoav. "Processing of natural images by the human visual system". Thesis, University of Cambridge, 1991. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.385855.
Texto completo
17
Heess, Nicolas Manfred Otto. "Learning generative models of mid-level structure in natural images". Thesis, University of Edinburgh, 2012. http://hdl.handle.net/1842/5866.
Texto completoResumen
Natural images arise from complicated processes involving many factors of variation. They reflect the wealth of shapes and appearances of objects in our three-dimensional world, but they are also affected by factors such as distortions due to perspective, occlusions, and illumination, giving rise to structure with regularities at many different levels. Prior knowledge about these regularities and suitable representations that allow efficient reasoning about the properties of a visual scene are important for many image processing and computer vision tasks. This thesis focuses on models of image structure at intermediate levels of complexity as required, for instance, for image inpainting or segmentation. It aims at developing generative, probabilistic models of this kind of structure, and, in particular, at devising strategies for learning such models in a largely unsupervised manner from data. One hallmark of natural images is that they can often be decomposed into regions with very different visual characteristics. The main approach of this thesis is therefore to represent images in terms of regions that are characterized by their shapes and appearances, and an image is then composed from many such regions. We explore approaches to learn about the appearance of regions, to learn about region shapes, and ways to combine several regions to form a full image. To achieve this goal, we make use of some ideas for unsupervised learning developed in the literature on models of low-level image structure and in the “deep learning” literature. These models are used as building blocks of more structured model formulations that incorporate additional prior knowledge of how images are formed. The thesis makes the following contributions: Firstly, we investigate a popular, MRF based prior of natural image structure, the Field-of Experts, with respect to its ability to model image textures, and propose an extended formulation that is considerably more successful at this task. This formulation gives rise to a fully parametric, translation-invariant probabilistic generative model of image textures. We illustrate how this model can be used as a component of a more comprehensive model of images comprising multiple textured regions. Secondly, we develop a model of region shape. This work is an extension of the “Masked Restricted Boltzmann Machine” proposed by Le Roux et al. (2011) and it allows explicit reasoning about the independent shapes and relative depths of occluding objects. We develop an inference and unsupervised learning scheme and demonstrate how this shape model, in combination with the masked RBM gives rise to a good model of natural image patches. Finally, we demonstrate how this model of region shape can be extended to model shapes in large images. The result is a generative model of large images which are formed by composition from many small, partially overlapping and occluding objects.
18
Summers, Robert James. "Human sensitivity to higher-order statistical structure in natural images". Thesis, University of Derby, 2005. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.418675.
Texto completo
19
Rickert, Thomas D. (Thomas Dale) 1975. "Texture-based statistical models for object detection in natural images". Thesis, Massachusetts Institute of Technology, 1999. http://hdl.handle.net/1721.1/80570.
Texto completoResumen
Thesis (S.B. and M.Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1999.Includes bibliographical references (p. 63-65).
by Thomas D. Rickert.
S.B.and M.Eng.
20
Sundin, Hannes y Jakob Josefsson. "Evaluating synthetic training data for character recognition in natural images". Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-280292.
Texto completoResumen
This thesis is centered around character recognition in natural images. More specifically, evaluating the use of synthetic font images for training a Convolutional Neural Network (CNN), compared to natural training data. Training a CNN to recognize characters in natural images often demands a large amount of labeled data. One alternative is to instead generate synthetic data by using digital fonts. A total of 41,664 font images were generated, which in combination with already existing data yielded around 99,000 images. Using this synthetic dataset, the CNN was trained by incrementally increasing synthetic training data and tested on natural images. At the same time, different preprocessing methods were applied to the synthetic data in order to observe the effect on accuracy. Results show that even when using the best performing pre-processing method and having access to 99,000 synthetic training images, a smaller set of natural training data yielded better results. However, results also show that synthetic data can perform better than natural data, provided that a good preprocessing method is used and if the supply of natural images is limited.I det här kandidatexamensarbetet behandlas bokstavigenkänning i naturliga bilder. Mer specifikt jämförs syntetiska typsnittsbilder med naturliga bilder för träning av ett Convolutional Neural Network (CNN). Att träna ett CNN för att känna igen bokstäver i naturliga bilder kräver oftast mycket betecknad naturlig data. Ett alternativ till detta är att producera syntetisk träningsdata i form av typsnittsbilder. I denna studie skapades 41664 typsnittsbilder, vilket i kombination med existerande data gav oss omkring 99 tusen syntetiska träningsbilder. Därefter tränades ett CNN med typsnittsbilder i ökande mängd för att sedan testas på naturliga bilder av bokstäver. Resultatet av detta jämfördes sedan med resultatet av att träna med naturliga bilder. Dessutom experimenterades med olika förbehandlingsmetoder för att observera förbehandlingens påverkan på klassifikationsgraden. Resultaten visade att även med den förbehandlingsmetoden som gav bäst resultat och med mycket mer data, var träning med syntetiska bilder inte lika effektivt som med naturliga bilder. Dock så visades det att med en bra förbehandlingsmetod kan syntetiska bilder ersätta naturliga bilder, givet att tillgången till naturliga bilder är begränsat.
21
Brifkany, Jan y Yasini Anass El. "Text Recognition in Natural Images : A study in Text Detection". Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-282935.
Texto completoResumen
In recent years, a surge in computer vision methods and solutions has been developed to solve the computer vision problem. By combining different methods from different areas of computer vision, computer scientists have been able to develop more advanced and sophisticated models to solve these problems. This report will cover two categories, text detection and text recognition. These areas will be defined, described, and analyzed in the result and discussion chapter. This report will cover an exciting and challenging topic, text recognition in natural images. It set out to assess the improvement of OCR accuracy after three image segmentation methods have been applied to images. The methods used are Maximally stable extremal regions and geometric filtering based on geometric properties. The result showed that the accuracy of OCR with segmentation methods had an overall better accuracy when compared to OCR without segmentation methods. Also, it was shown that images with horizontal text orientation had better accuracy when applying OCR with segmentation methods compared to images with multi-oriented text orientation.Under de senaste åren har en ökning av datorseende metoder och lösningar utvecklats för att lösa datorseende problemet. Genom att kombinera olika metoder från olika områden av datorseende har datavetare kunnat utveckla mer avancerade och komplexa modeller för att lösa dessa problem. Denna rapport kommer att omfatta två kategorier, textidentifiering och textigenkänning. Dessa områden kommer att definieras, beskrivas och analyseras i resultat- och diskussionskapitlet. Denna rapport kommer att omfatta ett mycket intressant och utmanande ämne, textigenkänning i naturliga bilder. Rapporten syftar till att bedöma förbättringen av OCR-resultatet efter det att tre bildsegmenteringsmetoder har tillämpats på bilder. Metoderna som har använts är ” Maximally stable extremal regions” och geometrisk filtrering baserad på geometriska egenskaper. Resultatet visade att hos OCR med segmenteringsmetoder hade en övergripande bättre resultat jämfört med OCR utan segmenteringsmetoder. Det visades också att bilder med horisontell textorientering hade bättre noggrannhet vid tillämpning av OCR med segmenteringsmetoder jämfört med bilder med flerorienterad textorientering.
22
Tian, Qi-Chong. "Color Correction and Contrast Enhancement for Natural Images and Videos". Thesis, Paris Sciences et Lettres (ComUE), 2018. http://www.theses.fr/2018PSLED020/document.
Texto completoResumen
L'amélioration d'image est une sorte de technique pour améliorer la qualité visuelle d'image, qui joue un rôle très important dans les domaines du traitement d'image et de la vision d'ordinateur. En particulier, nous considérons la correction de couleur et l'amélioration de contraste pour améliorer la qualité d'image.Dans la première partie de cette thèse, nous nous concentrons sur la correction des couleurs pour les images naturelles. Tout d'abord, nous donnons un examen simple de la correction des couleurs. Deuxièmement, nous proposons une méthode efficace de correction des couleurs pour la couture d'images via la spécification d'histogramme et la cartographie globale. Troisièmement, nous présentons une approche de cohérence des couleurs pour les collections d'images, basée sur la spécification de la gamme conservation histogramme.Dans la deuxième partie, nous prêtons attention à l'amélioration du contraste pour les images et les vidéos naturelles. Tout d'abord, nous donnons un simple examen de l'amélioration du contraste. Deuxièmement, nous proposons une méthode de préservation du contraste global de naturalité, qui peut éviter une survalorisation. Troisièmement, nous présentons une méthode de fusion à base de variation pour l'amélioration de l'image d'illumination non uniforme, qui peut éviter la sur-amplification ou la sous-amélioration. Enfin, nous étendons le cadre basé sur la fusion pour améliorer les vidéos avec une stratégie temporellement cohérente, qui n'entraîne pas de scintillement des artefactsImage enhancement is a kind of technique to improve the image visual quality, which plays a very important role in the domains of image processing and computer vision. Specifically, we consider color correction and contrast enhancement to improve the image quality.In the first part of this thesis, we focus on color correction for natural images. Firstly, we give a simple review of color correction. Secondly, we propose an efficient color correction method for image stitching via histogram specification and global mapping. Thirdly, we present a color consistency approach for image collections, based on range preserving histogram specification.In the second part, we pay attention to contrast enhancement for natural images and videos. Firstly, we give a simple review of contrast enhancement. Secondly, we propose a naturalness preservation global contrast enhancement method, which can avoid over-enhancement. Thirdly, we present a variational-based fusion method for non-uniform illumination image enhancement, which can avoid overenhancement or under-enhancement. Finally, we extend the fusion-based framework to enhance videos with a temporally consistent strategy, which does not result in flickering artifacts
23
Kraljevic, Matija. "Character recognition in natural images : Testing the accuracy of OCR and potential improvement by image segmentation". Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-187991.
Texto completoResumen
In recent years, reading text from natural images has gained renewed research attention. One of the main reasons for this is the rapid growth of camera-based applications on smart phones and other portable devices. With the increasing availability of high performance, low-priced, image-capturing devices, the application of scene text recognition is rapidly expanding and becoming increasingly popular. Despite many efforts, character recognition in natural images, is still considered a challenging and unresolved problem. The difficulties stem from the fact that natural images suffer from a wide variety of obstacles such as complex backgrounds, font variation, uneven illumination, resolution problems, occlusions, perspective effects, just to mention a few. This paper aims to test the accuracy of OCR in character recognition of natural images as well as testing the possible improvement in accuracy after implementing three different segmentation methods.The results showed that the accuracy of OCR was very poor and no improvments in accuracy were found after implementing the chosen segmentation methods.
24
Gironés, Sancho Xavier. "Real-Time Localization of Multi-Oriented Text in Natural Scene Images". Doctoral thesis, Universitat Rovira i Virgili, 2021. http://hdl.handle.net/10803/671518.
Texto completoResumen
Aquesta tesi es centra en el problema de la localització de textos en imatges d'escenes naturals des de la perspectiva de l'eficiència. Amb aquesta finalitat, s'introdueix un mètode de localització de text multiorientat en imatges naturals adequat per al processament en temps real de vídeo d'alta definició en dispositius portàtils i mòbils. El mètode proposat es basa en components connectats (CCs): Primer, els CCs s'aïllen convolucionant la imatge a múltiples escales amb un filtre espacial lineal dissenyat específicament, seguit de binarització amb histèresi. A continuació, els CCs no textuals es filtren emprant una cascada de classificadors locals que operen sobre descriptors expandits incrementalment, on la propietat d'ample de traç s'estima eficientment calculant els quadrats màxims inscrits en els CCs. Els CCs candidats i els seus veïns es verifiquen posteriorment amb un classificador amb context que té en compte els CC de destinació i el seu entorn. Finalment, s'extreuen seqüències de text de totes les escales i després es fusionen utilitzant programació dinàmica. El mètode proposat és capaç de processar vídeo HD de 1080p a gairebé 30 quadres per segon en un ordinador portàtil estàndard sense requerir GPU. A més, en comparatives realitzades en les bases de dades de lectura robusta de ICDAR 2013 i de text d'escena incidental de ICDAR 2015, la solució proposada va rendir més del doble de ràpid que l'estat de la tècnica, a banda de demostrar resultats competitius en termes de precisió i exhaustivitat.
A més, aquesta tesi introdueix una nova família d'aproximacions racionals de la funció arctangent vàlides en el rang [0, π / 2] que es pot ampliar fàcilment a dos i quatre quadrants, i una nova tècnica per a la localització de matrícules de vehicles en imatges naturals.Esta tesis se centra en el problema de la localización de textos en imágenes de escenas naturales desde la perspectiva de la eficiencia. Con este fin, se introduce un método de localización de texto multiorientado en imágenes naturales adecuado para el procesamiento en tiempo real de video de alta definición en dispositivos portátiles y móviles. El método propuesto se basa en el enfoque de componentes conectados (CCs): Primero, los CCs se aíslan convolucionando la imagen a múltiples escalas con un filtro espacial lineal diseñado específicamente, seguido de binarización con histéresis. A continuación, los CCs no textuales se filtran empleando una cascada de clasificadores locales que operan sobre descriptores expandidos incrementalmente, donde la propiedad de ancho de trazo se estima eficientemente calculando los cuadrados máximos inscritos en los CCs. Los CCs candidatos y sus vecinos se verifican posteriormente con un clasificador con contexto que tiene en cuenta los CC de destino y su entorno. Por último, se extraen secuencias de texto en todas las escalas y después se fusionan utilizando programación dinámica. El método propuesto es capaz de procesar video HD de 1080p a casi 30 cuadros por segundo en una computadora portátil estándar sin requerir una GPU. Además, en comparativas realizadas en las bases de datos de lectura robusta de ICDAR 2013 y de texto de escena incidental de ICDAR 2015, la solución propuesta desempeñó más del doble de rápido que el estado de la técnica, aparte de demostrar resultados competitivos en términos de precisión y exhaustividad. Además, esta tesis introduce una nueva familia de aproximaciones racionales de la función arcotangente válida en el rango [0, π / 2] que se puede ampliar fácilmente a dos y cuatro cuadrantes, y una nueva técnica para la localización de matrículas de vehículos en imágenes naturales.
This thesis focuses on the problem of text localization in natural scene images from the perspective of time-efficiency. Towards this end, a multi-oriented text localization method in natural images suitable for real-time processing of high-definition video on portable and mobile devices is introduced. The proposed method is based on the connected component (CC) approach: First, CCs are isolated by convolving a multi-scale pyramid with a specifically designed linear spatial filter, followed by hysteresis thresholding. Next, non-textual CCs are pruned employing a cascade of local classifiers fed with increasingly extended feature vectors, where the stroke width feature is estimated in linear time complexity by computing the maximal inscribed squares in the CCs. Candidate CCs and their neighbors are subsequently checked with a context-aware classifier that takes into account the target CCs and their vicinity. Lastly, text sequences are extracted in all pyramid levels and fused using dynamic programming. The proposed method is capable of processing 1080p HD video at nearly 30 frames per second on a standard laptop without requiring a GPU. Furthermore, when benchmarked on the ICDAR 2013 Robust Reading and on the ICDAR 2015 Incidental Scene Text datasets, it performed more than twice faster than the state-of-the-art, while still delivering competitive results in terms of precision and recall. Additionally, this thesis introduces a new family or rational approximations of the arctangent function valid in the [0, π/2] range that can be easily extended to two and four quadrants, and a new technique for vehicle license plate localization in unconstrained environments is presented as a practical use case leveraging the text localization system described in this research.
25
Lauritzen, Jan Soren. "Contrast normalisation and the visual coding of contrast in natural images". Thesis, University of Cambridge, 2005. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.615107.
Texto completo
26
Borikar, Siddharth Rajkumar. "FAST ALGORITHMS FOR FRAGMENT BASED COMPLETION IN IMAGES OF NATURAL SCENES". Master's thesis, University of Central Florida, 2004. http://digital.library.ucf.edu/cdm/ref/collection/ETD/id/4424.
Texto completoResumen
Textures are used widely in computer graphics to represent fine visual details and produce realistic looking images. Often it is necessary to remove some foreground object from the scene. Removal of the portion creates one or more holes in the texture image. These holes need to be filled to complete the image. Various methods like clone brush strokes and compositing processes are used to carry out this completion. User skill is required in such methods. Texture synthesis can also be used to complete regions where the texture is stationary or structured. Reconstructing methods can be used to fill in large-scale missing regions by interpolation. Inpainting is suitable for relatively small, smooth and non-textured regions. A number of other approaches focus on the edge and contour completion aspect of the problem. In this thesis we present a novel approach for addressing this image completion problem. Our approach focuses on image based completion, with no knowledge of the underlying scene. In natural images there is a strong horizontal orientation of texture/color distribution. We exploit this fact in our proposed algorithm to fill in missing regions from natural images. We follow the principle of figural familiarity and use the image as our training set to complete the image.M.S.
School of Computer Science
Engineering and Computer Science
Computer Science
27
Martin, Robert. "Contrast coding in natural images : a study of human eye movements". Thesis, University of Newcastle Upon Tyne, 2004. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.417516.
Texto completo
28
Tuke, C. E. "Content based semi-invariant search for natural, symbolic and sketch images". Thesis, University of York, 2004. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.422533.
Texto completo
29
Li, Wen-Hao. "Significance of multiple scattering in remotely sensed images of natural surfaces /". Thesis, Connect to this title online; UW restricted, 1997. http://hdl.handle.net/1773/6743.
Texto completo
30
Mattaparthi, Sai Venkata Akshay. "The Impact of Hexagonal grid on thePrincipal Component of Natural Images". Thesis, Blekinge Tekniska Högskola, Institutionen för tillämpad signalbehandling, 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:bth-17236.
Texto completoResumen
The visual processing in the real world is different from the digital world. Monkey’s and a human’s visual world is richer and more colourful affording sight of flies, regardless of whether they are immobile or airborne. The study of the evolutionary process of our visual system indicates the existence of variationally spatial arrangement; from densely hexagonal in the fovea to a sparse circular structure in the peripheral retina. Normally we use a rectangular grid for the processing of images. But as per the perspective of the human eyes, the new approach is to change the grid from rectangular to hexagonal. Applying hexagonal grid in image processing is very advantageous and easy for mimicking human visual system. The main advantages for using the hexagonal structure in image processing is its resemblance to the arrangement of photoreceptors in the human eyes. The visual processing in the real world is different from the digital world. Monkey’s and a human’s visual world is richer and more colourful affording sight of flies, regardless of whether they are immobile or airborne. The study of the evolutionary process of our visual system indicates the existence of variationally spatial arrangement; from densely hexagonal in the fovea to a sparse circular structure in the peripheral retina. Normally we use a rectangular grid for the processing of images. But as per the perspective of the human eyes, the new approach is to change the grid from rectangular to hexagonal. Applying hexagonal grid in image processing is very advantageous and easy for mimicking human visual system. The main advantages for using the hexagonal structure in image processing is its resemblance to the arrangement of photoreceptors in the human eyes.
31
Sarma, Subramonia P. "Relationship between suspicious coincidence in natural images and contour-salience in oriented filter responses". Thesis, Texas A&M University, 2003. http://hdl.handle.net/1969.1/472.
Texto completoResumen
Salient contour detection is an important lowlevel visual process in the human visual system, and has significance towards understanding higher visual and cognitive processes. Salience detection can be investigated by examining the visual cortical response to visual input. Visual response activity in the early stages of visual processing can be approximated by a sequence of convolutions of the input scene with the difference-of-Gaussian (DoG) and the oriented Gabor filters. The filtered responses are unusually high for prominent edge locations in the image, and are uniformly similar across different natural image inputs. Furthermore, such a response follows a power law distribution. The aim of this thesis is to examine how these response properties could be utilized to the problem of salience detection. First, I identify a method to find the best threshold on the response activity (orientation energy) toward the detection of salient contours: compare the response distribution to a Gaussian distribution of equal variance. Second, I justify this comparison by providing an explanation under the framework of Suspicious Coincidence proposed by Barlow [1]. A connection is provided between perceived salience of contours and the neuronal goal of detecting suspiciousness, where salient contours are seen as affording suspicious coincidences by the visual system. Finally, the neural plausibility of such a salience detection mechanism is investigated, and the representational effciency is shown which could potentially explain why the human visual system can effortlessly detect salience.
32
Fei, Xiao Lei. "Hybrid segmentation on slant & skewed deformation text in natural scene images". Thesis, University of Macau, 2010. http://umaclib3.umac.mo/record=b2182857.
Texto completo
33
Bodin, Emanuel. "Furniture swap : Segmentation and 3D rotation of natural images using deep learning". Thesis, Uppsala universitet, Signaler och system, 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-435503.
Texto completoResumen
Learning to perceive scenes and objects from 2D images as 3D models is atrivial task for a human but very challenging for a computer. Being ableto retrieve a 3D model from a scene just by taking a picture of it canbe of great use in many fields, for example when making 3D blueprintsfor buildings or working with animations in the game or film industry.Novel view synthesis is a field within deep learning where generativemodels are trained to construct 3D models of scenes or objects from 2Dimages. In this work, the generative model HoloGAN is combined together with aU-net segmentation network. The solution is able to, given an imagecontaining a single object as input, swap that object to another oneand then perform a rotation of the scene, generating new images fromunobserved view points. The segmentation network is trained with pairedsegmentation masks while HoloGAN is able to in an unsupervised mannerlearn 3D metrics of scenes from unlabeled 2D images. The system as awhole is trained on one dataset containing images of cars while theperformance of HoloGAN was evaluated on four additionaldatasets. The chosen method proved to be successful but came with somedrawbacks such as requiring large dataset sizes and being computationalexpensive to train.
34
Munn, Brandon Robert. "Critical Spatiotemporal Dynamics of the Brain’s Endogenous Activity and its Natural Stimuli". Thesis, The University of Sydney, 2019. https://hdl.handle.net/2123/21452.
Texto completoResumen
The endogenous activity and natural sensory input of the brain are rich in spatiotemporal dynamics. It has been hypothesised that the brain maintains an internal model that is optimally adapted to the statistical properties of its environment to facilitate efficient processing. However, the universal statistics of endogenous neural dynamics and spatiotemporal natural stimuli remain unclear. In this thesis, we investigate the collective statistical properties of natural time-varying images, individual neuronal spiking activity, and collective macroscale neural activity. We analyse an ensemble of natural time-varying images recorded using a neuromorphic camera from the perspective of their spatiotemporal patterns and find evidence of dynamical thermodynamic criticality with critical dynamics belonging to the universality class of directed percolation. We then analyse spiking activity in subcortical and cortical visual areas. We find fractal spiking is prevalent across the early visual system and that it is intrinsically linked to a neurons population coupling. We reproduce this spiking activity relationship using a fractal inhomogeneous Poisson process and establish a unified account of dynamic spiking properties in the visual pathway. We also analyse multielectrode cortical recordings and demonstrate the coexistence of neuronal avalanches and LFP patterns. Neuronal avalanches are scale-free bursts of neuronal activity separated by periods of quiescence that are a signature of criticality within the brain. The low-frequency LFP patterns consist of simple travelling waves and complex radial patterns, whereas, at high-frequency, the patterns consist of localised bursts of activity. We demonstrate a mechanism by which the evolution of the LFP patterns coordinates spikes to form scale-free neural activity. Thus, our results indicate that the brain maintains an internal model with rich critical dynamics that match that of its environment for efficient representations.
35
Müller, Andreas [Verfasser]. "Methods for Learning Structured Prediction in Semantic Segmentation of Natural Images / Andreas Müller". Bonn : Universitäts- und Landesbibliothek Bonn, 2014. http://d-nb.info/1060787075/34.
Texto completo
36
Müller, Andreas Christian [Verfasser]. "Methods for Learning Structured Prediction in Semantic Segmentation of Natural Images / Andreas Müller". Bonn : Universitäts- und Landesbibliothek Bonn, 2014. http://nbn-resolving.de/urn:nbn:de:hbz:5n-37289.
Texto completo
37
Khondkar, B. K. "A study of measured texture in images of natural scenes under varying illumination conditions". Thesis, Cranfield University, 1995. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.484252.
Texto completo
38
Leong, Chee Wee. "Modeling Synergistic Relationships Between Words and Images". Thesis, University of North Texas, 2012. https://digital.library.unt.edu/ark:/67531/metadc177223/.
Texto completoResumen
Texts and images provide alternative, yet orthogonal views of the same underlying cognitive concept. By uncovering synergistic, semantic relationships that exist between words and images, I am working to develop novel techniques that can help improve tasks in natural language processing, as well as effective models for text-to-image synthesis, image retrieval, and automatic image annotation. Specifically, in my dissertation, I will explore the interoperability of features between language and vision tasks. In the first part, I will show how it is possible to apply features generated using evidence gathered from text corpora to solve the image annotation problem in computer vision, without the use of any visual information. In the second part, I will address research in the reverse direction, and show how visual cues can be used to improve tasks in natural language processing. Importantly, I propose a novel metric to estimate the similarity of words by comparing the visual similarity of concepts invoked by these words, and show that it can be used further to advance the state-of-the-art methods that employ corpus-based and knowledge-based semantic similarity measures. Finally, I attempt to construct a joint semantic space connecting words with images, and synthesize an evaluation framework to quantify cross-modal semantic relationships that exist between arbitrary pairs of words and images. I study the effectiveness of unsupervised, corpus-based approaches to automatically derive the semantic relatedness between words and images, and perform empirical evaluations by measuring its correlation with human annotators.
39
O'Hare, Louise. "Some aspects of visual discomfort". Thesis, University of St Andrews, 2013. http://hdl.handle.net/10023/3551.
Texto completoResumen
Visual discomfort is the adverse sensations, such as headaches and eyestrain, encountered on viewing certain stimuli. These sensations can arise under certain viewing conditions, such as stereoscopic viewing and prolonged reading of text patterns. Also, discomfort can occur as a result of viewing stimuli with certain spatial properties, including stripes and filtered noise patterns of particular spatial frequency. This thesis is an exploration of the stimulus properties causing discomfort, within the framework of two theoretical explanations. Both of the explanations relate to the stimuli being difficult for the visual system to process. The first is concerned with discomfort being the result of inefficient neural processing. Neural activity requires energy to process information, and stimuli that demand a lot of energy to be processed might be uncomfortable. The second explanation revolves around uncomfortable stimuli not being effective in driving the accommodative (focussing) response. Accommodation relies on the stimulus as a cue to drive the response effectively - an uninformative cue might result in discomfort from an uncertain accommodative response. The following research investigates both these possibilities using a combination of psychophysical experimentation, questionnaire-based surveys on non-clinical populations, and computational modelling. The implications of the work for clinical populations are also discussed.
40
Kellmeyer, David. "Detection of highway warning signs in natural video images using color image processing and neural network techniques on a PC". Ohio : Ohio University, 1992. http://www.ohiolink.edu/etd/view.cgi?ohiou1172782930.
Texto completo
41
Schmidt, Uwe [Verfasser], Stefan [Akademischer Betreuer] Roth y Paolo [Akademischer Betreuer] Favaro. "Half-quadratic Inference and Learning for Natural Images / Uwe Schmidt ; Stefan Roth, Paolo Favaro". Darmstadt : Universitäts- und Landesbibliothek Darmstadt, 2017. http://d-nb.info/1128309998/34.
Texto completo
42
Marin-Franch, Ivan. "Information-theoretic analysis of trichromatic images of natural scenes under different phases of daylight". Thesis, University of Manchester, 2009. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.496257.
Texto completoResumen
Trichromatic images of natural scenes contain information about the reflecting surfaces in those scenes. Colour is a very important source of information for identifying surfaces and making judgements about them. Judgements about surfaces in a scene by the formation provided by their colour can be done only if such colours remain stable under changes in the illumination. This approxmiate constancy of colours is not straightforward, as the spectra of the light reflected towards the eye can change considerably with the spectrum of the illummation. Therefore, the three photoreceptors in the eye respond differently to the different spectra reflected by the same surface under different ihuminants. Nevertheless, for a given scene, patterns of receptor responses under different illuminants are intrinsically mterrelated, and the Strength of these interrelations allow for the generation of approximately illuminantinvariant surface-colour codings. The aim of this thesis was to quantify objectively how much trichromatic information is potentially available to the eye and how much of it is actually retrieved for natural scenes under changes in illumination.
43
Kennedy, Gina Mikala. "Detecting changes in the phase spectra of natural-images by the human visual system". Thesis, University of Cambridge, 1997. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.627158.
Texto completo
44
Bai, Junjie. "Efficient optimization for labeling problems with prior information: applications to natural and medical images". Diss., University of Iowa, 2016. https://ir.uiowa.edu/etd/3045.
Texto completoResumen
Labeling problem, due to its versatile modeling ability, is widely used in various image analysis tasks. In practice, certain prior information is often available to be embedded in the model to increase accuracy and robustness. However, it is not always straightforward to formulate the problem so that the prior information is correctly incorporated. It is even more challenging that the proposed model admits efficient algorithms to find globally optimal solution.
In this dissertation, a series of natural and medical image segmentation tasks are modeled as labeling problems. Each proposed model incorporates different useful prior information. These prior information includes ordering constraints between certain labels, soft user input enforcement, multi-scale context between over-segmented regions and original voxel, multi-modality context prior, location context between multiple modalities, star-shape prior, and gradient vector flow shape prior.
With judicious exploitation of each problem's intricate structure, efficient and exact algorithms are designed for all proposed models. The efficient computation allow the proposed models to be applied on large natural and medical image datasets using small memory footprint and reasonable time assumption. The global optimality guarantee makes the methods robust to local noise and easy to debug.
The proposed models and algorithms are validated on multiple experiments, using both natural and medical images. Promising and competitive results are shown when compared to state-of-art.
45
Shaw, Lynda Joan. "Emotional processing of natural visual images in brief exposures and compound stimuli : fMRI and behavioural studies". Thesis, Brunel University, 2009. http://bura.brunel.ac.uk/handle/2438/3203.
Texto completoResumen
Can the brain register the emotional valence of brief exposures of complex natural stimuli under conditions of forward and backward masking, and under conditions of attentional competition between foveal and peripheral stimuli? To address this question, three experiments were conducted. The first, a behavioural experiment, measured subjective valence of response (pleasant vs unpleasant) to test the perception of the valence of natural images in brief, masked exposures in a forward and backward masking paradigm. Images were chosen from the International Affective Picture System (IAPS) series. After correction for response bias, responses to the majority of target stimuli were concordant with the IAPS ratings at better than chance, even when the presence of the target was undetected. Using functional magnetic resonance imaging (fMRI), the effects of IAPS valence and stimulus category were objectively measured on nine regions of interest (ROIs) using the same strict temporal restrictions in a similar masking design. Evidence of affective processing close to or below conscious threshold was apparent in some of the ROIs. To further this line of enquiry, a second fMRI experiment mapping the same ROIs and using the same stimuli were presented in a foveal (‘attended’) peripheral (‘to-be-ignored’) paradigm (small image superimposed in the centre of a large image of the same category, but opposite valence) to investigate spatial parameters and limitations of attention. Results are interpreted as showing both valence and category specific effects of ‘to-be-ignored’ images in the periphery. These results are discussed in light of theories of the limitations of attentional capacity and the speed in which we process natural images, providing new evidence of the breadth of variety in the types of affective visual stimuli we are able to process close to the threshold of conscious perception.
46
Wells, James Edward II. "Western landscapes, western images: a rephotography of U.S. Highway 89". Diss., Kansas State University, 2012. http://hdl.handle.net/2097/13524.
Texto completoResumen
Doctor of PhilosophyDepartment of Geography
Kevin Blake
The American West is a land of great diversity and stark contrast. It is also a landscape marked by rapid change as a result of such forces as globalization, population growth, and heightened interest in natural resources (either for recreation or extraction). This dissertation investigates these changes to the region through a repeat photography analysis. Between 1982 and 1984, Thomas and Geraldine Vale traveled along U.S. Highway 89 from Glacier National Park, Montana to Nogales, Arizona. Their subsequent work, Western Images, Western Landscapes: Travels Along U.S. 89 (University of Arizona Press, 1989), contained fifty-three photographs from this journey, representing a cross section of the West from border to border. Nearly every facet of the region was represented, from the remote prairie landscapes of Montana to the bustling Phoenix downtown, and from the largest open pit mine in the world to seldom visited corners of Yellowstone National Park. Between March 2009 and August 2010, I retraced the steps taken by the Vales and successfully rephotographed all of the locations contained within their book. The observed continuity or change is examined thematically in order to address the landscapes and cultures of the West in greater detail. Specifically, chapters within this dissertation visually and textually describe changes that have occurred along national borders, within Native American reservations, throughout the rural landscapes and national parks of the region, within the many resource extraction industries, and within towns and cities of every size. Significant findings, which are well depicted in the photographic pairings, include heightened national security along the borders, problems of overuse in many parks and protected areas, the transition of traditional small towns into communities increasingly dependent upon tourism for survival, and both beautification and revitalization efforts taking place in the urban cores of Phoenix and Salt Lake City. By painting a vivid picture of recent Western geography, this research provides for greater ability for residents and scholars of the region to understand the forces at work within their communities and surroundings.
47
Chen, Ying-Yu y 陳盈瑜. "Secret Image Sharing with Natural Shadow Images". Thesis, 2009. http://ndltd.ncl.edu.tw/handle/86133119357647877403.
Texto completoResumen
碩士銘傳大學
資訊傳播工程學系碩士班
97
Secret image sharing is one of the protection methods for important images. In the traditional (r, n)-threshold image sharing scheme, the important image is encoded into n noise-like shadow images in such a way that the original image can be revealed if r or more of these n shadows are obtained, but less than r shadow images cannot get anything about the secret image. The main characteristic of this technology is dispersing the important image in several smaller shadow images, which benefits the transmission and storage of the shadow images. The method also has a high fault-tolerance against data incomplete in that the original image can be reconstructed when fewer than n-r of the shadow images are damaged or lost. This study proposes a secret image sharing method with natural shadow images. The concept is by comparing the values of a shadow key and the generated shadow pixel with the pixel of the cover images. The pixel is replaced with the most similar one among all of the candidates. It makes the generated images look naturally, and attracts less attention to unintended users. From the experimental results, we see that the quality of shadow images obtained by the proposed method is acceptable. In addition, the secret key is also shared and stored in the shadow images without extra storage.
48
Tsang, Chien-Chu y 臧建柱. "Sharing Secret Images by Natural Images". Thesis, 2011. http://ndltd.ncl.edu.tw/handle/90298639326569617805.
Texto completoResumen
碩士銘傳大學
資訊工程學系碩士班
99
Conventional visual secret sharing (VSS) schemes generate noise-like random pixels on shares to hide secret images. However, it suffers two problems. First, it has a management problem, because of which participants cannot visually identify each share, especially when they hold more than one share simultaneously. Second, the noise-like shares are easy to arouse suspicion that leads security problems for participants. Previous researchers developed extended visual cryptography schemes that add a meaningful cover image on each share cope with the management problem. Generally, these approaches introduce a more serious pixel expansion problem than conventional VSS schemes. Moreover, there are still has many noise-like pixels on the shares that cannot effectively reduce the security problems. This paper proposes 3 new secret sharing schemes: (n, n), (n, n) multiple secret sharing, and (k, n) scheme. All the above schemes can share a binary secret image over n-1 arbitrary natural images and one noise-like share image. Instead of altering the contents of these natural images, the encryption process extracts feature images from each natural image. In such a way, these unaltered natural images are totally innocuous; the security problem can be greatly reduced. Upon the natural images and the share image received, the feature images can be extracted and the secret image can be recovered by cooperating these feature images and the share. The experimental results indicate that the proposed approach can provide an excellent solution for solving the management and security problems. Moreover, the proposed approach can avoid the pixel expansion problem.
49
Li, Yu-Chai y 李雨儕. "Symbol of Natural Images". Thesis, 2010. http://ndltd.ncl.edu.tw/handle/ccxhxb.
Texto completoResumen
碩士國立臺中教育大學
美術學系碩士班
99
The human society is part of nature. Human and the Nature are in a state of eternally interactive relationship. We have not been separated from the Nature since the day we were born. We gradually come into contact with extrinsic perception and develop intrinsic heart-throb. The creation course is a subjective soul transformation blended with individual subjectivity and media of objectivity, or it even forms mutual repulse or confrontation. Particularly in the process of creation, the mindset and performing style of the creator forming a mutual connection and tug, is regarded as a pleasant course with loneliness and painful experiences in the process of art creation. The author has been quite fortunate in the course of learning painting. He applies the creating experience of forerunners in the learning process and reads Chinese, foreign, historical, and modern painting theories and aesthetics philosophy, to clarify his myth in creation so that the creation process is no longer restrained by media application, increasing the breadth and depth of self artistic language and thoughts of creations. His ideas and context will diversify his creation styles and assists him in reaching higher level of creations. The study divides the creation course and research orientation of the author into three sections: 1. Discuss of the Views on Glue Color Painting from historical context. The study conducts in-depth research on the Taiwanese painters during the Japanese Colonial Period under the influence of Japanese painting through the development of contemporary Japanese paintings, in order to understand the views on glue color painting upon the first introduction to Taiwanese painting industry. Taiwan’s glue color painting underwent difficult heritage from a politically correct era and with national complex and change of time, and reaches a thriving development today. The study discusses the development course of views on glue color painting and makes an analysis and comparison for the influence on the author’s technique in glue color painting. 2. Analysis the formation of individual creation concepts through Oriental and Western philosophies The study analyzes the concept of individual art creation through the nature philosophy based on Oriental Confucian, Daoism and Buddhism and the Western aesthetics ideas of nature. Moreover, the creation content and ideas based on nature will undergo statement and analysis to induce elements of self-painting styles thereby used as direction and reference for future creations. 3. Analysis and Statement of Creation Course The author makes reflection on the creation course based on “Symbol of Natural Images” and the “Nature of Arts and Culture” of art creation under the influence of avant-garde artistic ideas in post-modernism, localization and globalization. The author expects to add more cultural traits in his own creation in the future, expanding the context and possibility of self-creation.
50
Ahmed, Faruk. "Generative models for natural images". Thèse, 2017. http://hdl.handle.net/1866/20186.
Texto completo