No description available.

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2015-05-26T17:01:53Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Dissertação Mestrado Arthur de Lima Padilha 2015.pdf: 4117834 bytes, checksum: f6a934e1c3d8dfce94e7122fc51183d0 (MD5) / Made available in DSpace on 2015-05-26T17:01:53Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Dissertação Mestrado Arthur de Lima Padilha 2015.pdf: 4117834 bytes, checksum: f6a934e1c3d8dfce94e7122fc51183d0 (MD5) Previous issue date: 2015-03-13 / Em Realidade Aumentada funcional o objetivo principal é ajudar os observadores com a execução de tarefas através de uma percepção melhorada da cena aumentada. Este trabalho aplica uma técnica de visualização bastante utilizada, conhecida como Ghosting, para melhorar a percepção de profundidade em cenas de Realidade Aumentada. Técnicas de visão computacional e processamento de imagens são usadas para extrair as características naturais em cenas reais para, em seguida, decidir a quantidade de transparência que será aplicada em cada pixel do objeto virtual, e assim, gerar o efeito de ghosting quando misturar o objeto virtual na cena real. Um objeto em movimento em uma cena real chama bastante atenção do observador. Assim, espera-se que, quando um objeto em movimento passar na frente de um objeto aumentado, a informação visual do objeto real se mantenha, para que a visualização de informações naturais, e importantes da cena, não sejam perdidas. Por isso, a principal contribuição deste trabalho é a inclusão de uma técnica de detecção de movimento na etapa de análise de características da cena do pipeline da técnica de Ghosting. A avaliação qualitativa realizada com diversos usuários mostrou que os estudos de caso, em ambientes internos e externos, abordados neste trabalho, proporcionaram uma melhor percepção de profundidade da cena de Realidade Aumentada, pois preserva as informações mais relevantes para a correta percepção do ambiente pelo usuário. / Functional Augmented Reality realism focuses on helping users with tasks execution through an enhanced perception of the augmented scene. This work applies a common visualization technique, Ghosting, to improve depth perception in Augmented Reality scenes. Computer Vision and Image Processing techniques are used to extract natural features from a real scene, which will guide the assignment of transparency to each pixel of the virtual object, and provide the ghosting effect while blending the virtual object into the real scene. A moving objet in a real scene catches user attention. So, it's expected that natural and important visual information of the scene don't get occluded when the moving object passes over it. Because of that, the main contribution of this work is the inclusion of a motion detection technique to the scene feature analysis step, into the Ghosting technique pipeline. The qualitative evaluation showed that the user's studies of this work, in indoor and outdoor environments, using the proposed technique leaded to a better depth perception of the augmented scene, because it preserves the most relevant information for correct user's perception of the scene.

Tato diplomová práce se zabývá syntézou HDR obrazu (High Dynamic Range Imaging). HDRI technologie se stala v posledních letech velice populární. Běžný a nejvíce používaný způsob vytvoření HDR obrazu je spojení více snímků stejné scény pořízených pomocí různých expozičních časů. Tato technika funguje správně pouze v případě, že se jedná o statickou scénu. Pokud je však ve scéně nějaký pohyb ve chvíli, kdy se pořizují snímky dané scény, výsledný HDR obraz obsahuje artefakty zvané jako duchy. V této práci jsou prezentovány základní informace o HDRI se zaměřením na metody odstraňující artefakty z HDR obrazů. Práce shrnuje již existující metody a dvě z nich - tzv. bitmap movement detection a histogram based ghost detection - představuje jako vhodné pro použití v real-time skládání HDR obrazu a pro implementaci na FPGA (Field-Programmable Gate Array) architektuře. Tyto metody jsou v práci implementovány v programovacím jazyce C++ jako prototypy. Navíc je zde navržena modifikace metody založené na výpočtu histogramu pro jednodušší a efektivnější implementaci na FPGA architektuře.

<p>Direct conversion flat panel x-ray image sensors based on using a photoconductor with an active matrix array provide excellent images. These image sensors are suitable for replacing the present day x-ray film/screen cassette to capture an x-ray image electronically, and hence enable a clinical transition to digital radiography. The performance of these sensors depends critically on the selection and design of the photoconductor. This work quantitatively studies the combined effects of the detector geometry (pixel size and detector thickness), operating conditions (x-ray energy and applied electric field) and charge transport properties (e.g., carrier trapping and recombination) of the photoconductor on the detector performance by developing appropriate detector models. In this thesis, the models for calculating the x-ray sensitivity, resolution in terms of the modulation transfer function (MTF), detective quantum efficiency (DQE), and ghosting of x-ray image detectors have been developed. The modeling works are based on the physics of the individual phenomena and the systematic solution of the fundamental physical equations in the photoconductor layer: (1) semiconductor continuity equation (2) Poissons equation (3) trapping rate equations. The general approach of this work is to develop models in normalized coordinates to describe the results of different photoconductive x-ray image detectors. These models are applied to a-Se, polycrystalline HgI_2 and polycrystalline CdZnTe photoconductive detectors for diagnostic medical x-ray imaging applications (e,g., mammography, chest radiography and fluoroscopy). The models show a very good agreement with the experimental results.</p><p>The research presented in this thesis shows that the imaging performances (e.g., sensitivity, MTF, DQE and ghosting) can be improved by insuring that the carrier with higher mobility-lifetime product is drifted towards the pixel electrodes. The carrier schubwegs have to be several times greater, and the absorption depth has to be at least two times smaller than the photoconductor thickness for achieving sufficient sensitivity. Having smaller pixels is advantageous in terms of higher sensitivity by ensuring that the carrier with the higher mobility-lifetime product is drifted towards the pixel electrodes. </p><p>A model for calculating zero spatial frequency detective quantum efficiency, DQE (0), has been developed by including incomplete charge collection and x-ray interaction depth dependent conversion gain. The DQE(0) analyses of a-Se detectors for fluoroscopic applications show that there is an optimum photoconductor thickness, which maximizes the DQE(0) under a constant voltage operation. The application of DQE(0) model to different potential photoconductive detectors for fluoroscopic applications show that, in addition to high quantum efficiency, both high conversion gain and high charge collection efficiency are required to improve the DQE performance of an x-ray image detector.</p><p>An analytical expression of MTF due to distributed carrier trapping in the bulk of the photoconductor has been derived using the trapped charge distribution across the photoconductor. Trapping of the carriers that move towards the pixel electrodes degrades the MTF performance, whereas trapping of the other type of carriers improves the sharpness of the x-ray image.</p><p>The large signal model calculations in this thesis show an upper limit of small signal models of x-ray image detectors. The bimolecular recombination between drifting carriers plays practically no role on charge collection in a-Se detectors up to the total carrier generation rate q0 of 10^18 EHPs/m^2-s. The bimolecular recombination has practically no effect on charge collection in a-Se detectors for diagnostic medical x-ray imaging applications. </p><p>A model for examining the sensitivity fluctuation mechanisms in a-Se detectors has been developed. The comparison of the model with the experimental data reveals that the recombination between trapped and the oppositely charged drifting carriers, electric field dependent charge carrier generation and x-ray induced new deep trap centers are mainly responsible for the sensitivity fluctuation in biased a-Se x-ray detectors. </p><p>The modeling works in this thesis identify the important factors that limit the detector performance, which can ultimately lead to the reduction of patient exposure/dose consistent with better diagnosis for different diagnostic medical x-ray imaging modalities. The quantitative analyses presented in this thesis show that the detector structure is just as important to the overall performance of the detector as the material properties of the photoconductor itself.</p>

<p>Direct conversion flat panel x-ray image sensors based on using a photoconductor with an active matrix array provide excellent images. These image sensors are suitable for replacing the present day x-ray film/screen cassette to capture an x-ray image electronically, and hence enable a clinical transition to digital radiography. The performance of these sensors depends critically on the selection and design of the photoconductor. This work quantitatively studies the combined effects of the detector geometry (pixel size and detector thickness), operating conditions (x-ray energy and applied electric field) and charge transport properties (e.g., carrier trapping and recombination) of the photoconductor on the detector performance by developing appropriate detector models. In this thesis, the models for calculating the x-ray sensitivity, resolution in terms of the modulation transfer function (MTF), detective quantum efficiency (DQE), and ghosting of x-ray image detectors have been developed. The modeling works are based on the physics of the individual phenomena and the systematic solution of the fundamental physical equations in the photoconductor layer: (1) semiconductor continuity equation (2) Poissons equation (3) trapping rate equations. The general approach of this work is to develop models in normalized coordinates to describe the results of different photoconductive x-ray image detectors. These models are applied to a-Se, polycrystalline HgI_2 and polycrystalline CdZnTe photoconductive detectors for diagnostic medical x-ray imaging applications (e,g., mammography, chest radiography and fluoroscopy). The models show a very good agreement with the experimental results.</p><p>The research presented in this thesis shows that the imaging performances (e.g., sensitivity, MTF, DQE and ghosting) can be improved by insuring that the carrier with higher mobility-lifetime product is drifted towards the pixel electrodes. The carrier schubwegs have to be several times greater, and the absorption depth has to be at least two times smaller than the photoconductor thickness for achieving sufficient sensitivity. Having smaller pixels is advantageous in terms of higher sensitivity by ensuring that the carrier with the higher mobility-lifetime product is drifted towards the pixel electrodes. </p><p>A model for calculating zero spatial frequency detective quantum efficiency, DQE (0), has been developed by including incomplete charge collection and x-ray interaction depth dependent conversion gain. The DQE(0) analyses of a-Se detectors for fluoroscopic applications show that there is an optimum photoconductor thickness, which maximizes the DQE(0) under a constant voltage operation. The application of DQE(0) model to different potential photoconductive detectors for fluoroscopic applications show that, in addition to high quantum efficiency, both high conversion gain and high charge collection efficiency are required to improve the DQE performance of an x-ray image detector.</p><p>An analytical expression of MTF due to distributed carrier trapping in the bulk of the photoconductor has been derived using the trapped charge distribution across the photoconductor. Trapping of the carriers that move towards the pixel electrodes degrades the MTF performance, whereas trapping of the other type of carriers improves the sharpness of the x-ray image.</p><p>The large signal model calculations in this thesis show an upper limit of small signal models of x-ray image detectors. The bimolecular recombination between drifting carriers plays practically no role on charge collection in a-Se detectors up to the total carrier generation rate q0 of 10^18 EHPs/m^2-s. The bimolecular recombination has practically no effect on charge collection in a-Se detectors for diagnostic medical x-ray imaging applications. </p><p>A model for examining the sensitivity fluctuation mechanisms in a-Se detectors has been developed. The comparison of the model with the experimental data reveals that the recombination between trapped and the oppositely charged drifting carriers, electric field dependent charge carrier generation and x-ray induced new deep trap centers are mainly responsible for the sensitivity fluctuation in biased a-Se x-ray detectors. </p><p>The modeling works in this thesis identify the important factors that limit the detector performance, which can ultimately lead to the reduction of patient exposure/dose consistent with better diagnosis for different diagnostic medical x-ray imaging modalities. The quantitative analyses presented in this thesis show that the detector structure is just as important to the overall performance of the detector as the material properties of the photoconductor itself.</p>

In this thesis the main objectives were to improve the liquid crystal shutters and to develop a stereoscopic vision application called 3D Dual View. The vision quality of a 3D system based on active glasses is direct related to the performance of the liquid crystal shutters. There are a lot of parameters that plays an important role in the quality of 3D glasses. Thus, efforts were concentrated in the organization of the liquid crystal molecules between the substrates and in the liquid crystal itself. We studied different configurations of shutters using nematic liquid crystal and ferroelectric liquid crystal (smectic), with and without polymer. Furthermore, the ferroelectric samples were analyzed using a full optical snapshot matrix Mueller polarimenter made by UBO (Université de Bretagne Occidentale) in partnership with Télécom-Bretagne. The second part was the development of 3D Dual View System. It is a stereoscopic vision system based on a mixture of active and passive 3D solution which displays a 3D image that can be viewed simultaneously by two spectators, but from different perspectives. To allow two different points-of-view of a scene or object in three dimensions, the system has to multiplex four images, one pair to form the 3D image for one viewer in one perspective and the other pair for the second viewer in other perspective. We describe different techniques to multiplex and demultiplex the video streams taking into account the physiological aspects, market, image quality and the crosstalk. The crosstalk between the images leads to an effect called ghosting, which the viewer sees a low intensity image that belongs to the other viewer. Then, we also characterized the ghosting effect (crosstalk) and proposed an algorithm of ghostbusting to compensate it.

This study researches communication 2020-2021 via the Tinder dating app in Sweden during the COVID-19 pandemic, using a framework of mediatization, uses and gratification theory and affordance theory. Data collection was performed using qualitative interviews via Zoom video calls with respondents based in Sweden and findings were analyzed using thematic analysis.  Results of the study discusses phenomena when the physical world and a mediatized one interlock: even though the Tinder app afforded people to connect during the COVID-19 pandemic, the worlds are not independent of each other. Uncertainty and stress could seep into the dating app practice and create a hindrance to easy communication, even though the Tinder app was trying to accommodate the changing times. The study explores that the dehumanizing aspect of reducing a person to a single attribute and deleting that person is connected to a heightened affordance of gamification of relationships that is connected to deep mediatization, moments of our everyday practices immersed in new kinds of mediated communication. Steered behaviour from the app occurs in the Tinder interface which this paper explores, connected to data-driven decision making, a critical tool allowing mobile app developers to not only figure out customer behaviour but also to design customised products. Connected here is the mediatized interplay between humans and technology. If the Tinder app was hard to set up or navigate, the different needs and gratifications would not be met and users would simply vanish.  This paper argues the Tinder app design encourages speed and has coded distractions to actually work against the perceived gratification of meeting a partner and ultimately rendering the service redundant. The COVID-19 pandemic, this thesis explores, has spurred different generations to adapt to new practices and technologies but that these affordances do not always correlate to new dating app practices such as video dating. A mediatized way of meeting physically has been brought on by the COVID-19 pandemic. An additional step has been added to online dating practices, the walk, which has 2 lowered the threshold of the initial date by being perceived as an easier option. A direct consequence of the COVID-19 pandemic connected to online dating. This paper also argues intimacy through media sharing, such as humour and jokes via technology in a mediatized world. Lastly, this paper explores the safety aspects of the Tinder app, connected to catfishing and also gender blindness.