<p>As real time visualizations become more realistic it also becomes more important to simulate the perceptual effects of the human visual system. Such effects include the response to varying illumination, glare and differences between photopic and scotopic vision. This thesis evaluates several different tone mapping methods to allow a greater dynamic range to be used in real time visualisations. Several tone mapping methods have been implemented in the Avalanche Game Engine and evaluated using a small test group. To increase immersion in the visualization several filters aimed to simulate perceptual effects has also been implemented. The primary goal of these filters is to simulate scotopic vision. The tests showed that two tone mapping methods would be suitable for the environment used in the tests. The S-curve tone mapping method gave the best result while the Mean Value method gave good results while being the simplest to implement and the cheapest. The test subjects agreed that the simulation of scotopic vision enhanced the immersion in a visualization. The primary difficulties in this work has been lack of dynamic range in the input images and the challenges in coding real time graphics using a graphics processing unit.</p>

The first chapter explains the motivation for measuring rod function, in particular the rod’s dynamic recovery from a substantial bleach which results in so-called ‘rate limited’ recovery of sensitivity. The physiological processes that underpin the replenishment of the rod photopigment are described and discussed, and explain the way in which rod function can act as a marker for retinal health. Overall, this chapter explains why rod function is worthy of further investigation.Then follows a description of the experimental methods used in the study of rod function, presented in later chapters. The psychophysical procedures are described and a new method of dark adaptation measurement is presented. The key feature of this technique is a red background.Nonlinear mathematical models are used to describe the reduction in visual thresholds with time following a bleach. Chapter three describes the difficulties associated with numerical methods of nonlinear regression and presents a novel, multi start algorithm that extracts the parameters of interest from a model that adequately describes dark adaptation in the healthy normal subject.Chapter 4 verifies the algorithm presented in chapter 3, which is shown to be reliable and robust. A series of numerical experiments are performed to evaluate some of the characteristics of the algorithm’s performance.In chapter five, a series of experiments are presented to investigate the possible effect of a luminous background on dark adaptation (DA). The first experiment tests whether the rod system can detect a dim red background and the second, whether the rod thresh olds, when measured against light emitted by a red light emitting diode (LED), were linear. The third explores whether the background had any effect on the recovery of rod sensitivity. Finally, conventional contrast sensitivity is used to investigate the recovery from a photo bleach.A novel laboratory based apparatus was used to measure dark adaptation in a group of 36 subjects and the results of these measurements are presented in chapter six. The aim here was to see if the data collected were comparable with the dark adaptation data in the literature. These subjects were asked to make two visits so that an assessment of the test retest reliability of the method could be made. The method is shown to be reliable and capable of characterising the recovery of the visual system after a photo bleach.Although inherently flexible the analogue apparatus was prone to subject driven variability. Greater consistency of measurement was achieved using a digital device developed in partnership with an industry partner, Elektron (UK). This device, described in chapter seven provided fine control of many of the experimental parameters. It was used to measure the dark adaptation of a young healthy group of 21 people.This study uses new methodological approaches, both experimental and statistical, that are robust and reliable to facilitate investigation of rod function, and presents new findings about the early phase of rod sensitivity recovery.

Recentes estudos sugerem a existência de sinapses elétricas mediadas por junções gap entre fotorreceptores na retina de vertebrados. Neste trabalho, descrevemos um modelo computacional dos circuitos primário e secundário mediados pelos bastonetes da retina de vertebrados. O modelo é composto pelas seguintes populações de células: bastonetes, cones, células bipolares dos bastonetes, células bipolares dos cones, células amácrinas do tipo AII e células ganglionares. As células do modelo estão acopladas entre si por sinapses químicas e elétricas segundo padrões realísticos de convergência e divergência. As sinapses elétricas ocorrem entre os bastonetes, entre os bastonetes e os cones, entre as células amácrinas AII e entre as células bipolares dos cones e a células amácrinas AII. O modelo assume que um estímulo luminoso de baixa intensidade, simulando condições escotópicas, atinge todos os bastonetes da camada receptora, porém menos da metade deles é excitada. A resposta dos bastonetes excitados é controlada por uma fotocorrente cuja amplitude pode ser alterada para simular estímulos de diferentes intensidades dentro da faixa escotópica. O modelo é utilizado para investigar os efeitos dos diferentes graus de acoplamento elétrico entre as células receptoras e entre as células amácrinas AII, além do efeito de diferentes valores de condutância do canal Ih ativado pela hiperpolarização nos bastonetes, sobre a faixa dinâmica da retina. Os resultados das simulações mostram que, para valores realísticos da condutância do canal Ih, a faixa dinâmica medida na camada receptora é maximizada para o índice de conectividade crítico para que haja percolação de ligação. No entanto, quando a faixa dinâmica é medida para as células bipolares ou ganglionares o valor máximo é obtido para um índice de conectividade subcrítico. Este resultado é conseqüência da alta convergência de sinapses químicas entre os bastonetes e células bipolares. / Recent studies suggest the existence of electrical synapses (gap junctions) connecting photoreceptors in the vertebrate retina. In this work we describe a computer model of the primary and secondary rod pathways in the vertebrate retina. The model is composed of the following cell populations: rods, cones, rod bipolar cells, cone bipolar cells, AII amacrine cells and ganglion cells. Cells of the model are connected via chemical as well as electrical synapses according to realistic convergence and divergence factors. There are electrical synapses between rods, rods and cones, AII amacrine cells, and cone bipolar cells and AII amacrine cells. The model assumes that low intensity stimuli simulating scotopic conditions reach all rods in the receptor array but less than half of them are excited. The excited rods response is controlled by a photocurrent waveform whose amplitude can be manipulated to simulate stimuli of different intensities within the scotopic range. The model is used to investigate the effects of different degrees of coupling among photoreceptors and among AII amacrine cells, as well as values of rod hyperpolarization activated current Ih on the dynamic range of the retina. Results show that for realistic values of Ih conductance the dynamic range of the rod array is maximized at the critical connectivity degree for bond percolation. However, the dynamic range of the rod bipolar and ganglion cells is maximized for a photoreceptor connectivity degree below the critical value. The latter result is a consequence of the high convergence of chemical synapses from rods to rod bipolar cells.

Recentes estudos sugerem a existência de sinapses elétricas mediadas por junções gap entre fotorreceptores na retina de vertebrados. Neste trabalho, descrevemos um modelo computacional dos circuitos primário e secundário mediados pelos bastonetes da retina de vertebrados. O modelo é composto pelas seguintes populações de células: bastonetes, cones, células bipolares dos bastonetes, células bipolares dos cones, células amácrinas do tipo AII e células ganglionares. As células do modelo estão acopladas entre si por sinapses químicas e elétricas segundo padrões realísticos de convergência e divergência. As sinapses elétricas ocorrem entre os bastonetes, entre os bastonetes e os cones, entre as células amácrinas AII e entre as células bipolares dos cones e a células amácrinas AII. O modelo assume que um estímulo luminoso de baixa intensidade, simulando condições escotópicas, atinge todos os bastonetes da camada receptora, porém menos da metade deles é excitada. A resposta dos bastonetes excitados é controlada por uma fotocorrente cuja amplitude pode ser alterada para simular estímulos de diferentes intensidades dentro da faixa escotópica. O modelo é utilizado para investigar os efeitos dos diferentes graus de acoplamento elétrico entre as células receptoras e entre as células amácrinas AII, além do efeito de diferentes valores de condutância do canal Ih ativado pela hiperpolarização nos bastonetes, sobre a faixa dinâmica da retina. Os resultados das simulações mostram que, para valores realísticos da condutância do canal Ih, a faixa dinâmica medida na camada receptora é maximizada para o índice de conectividade crítico para que haja percolação de ligação. No entanto, quando a faixa dinâmica é medida para as células bipolares ou ganglionares o valor máximo é obtido para um índice de conectividade subcrítico. Este resultado é conseqüência da alta convergência de sinapses químicas entre os bastonetes e células bipolares. / Recent studies suggest the existence of electrical synapses (gap junctions) connecting photoreceptors in the vertebrate retina. In this work we describe a computer model of the primary and secondary rod pathways in the vertebrate retina. The model is composed of the following cell populations: rods, cones, rod bipolar cells, cone bipolar cells, AII amacrine cells and ganglion cells. Cells of the model are connected via chemical as well as electrical synapses according to realistic convergence and divergence factors. There are electrical synapses between rods, rods and cones, AII amacrine cells, and cone bipolar cells and AII amacrine cells. The model assumes that low intensity stimuli simulating scotopic conditions reach all rods in the receptor array but less than half of them are excited. The excited rods response is controlled by a photocurrent waveform whose amplitude can be manipulated to simulate stimuli of different intensities within the scotopic range. The model is used to investigate the effects of different degrees of coupling among photoreceptors and among AII amacrine cells, as well as values of rod hyperpolarization activated current Ih on the dynamic range of the retina. Results show that for realistic values of Ih conductance the dynamic range of the rod array is maximized at the critical connectivity degree for bond percolation. However, the dynamic range of the rod bipolar and ganglion cells is maximized for a photoreceptor connectivity degree below the critical value. The latter result is a consequence of the high convergence of chemical synapses from rods to rod bipolar cells.

As real time visualizations become more realistic it also becomes more important to simulate the perceptual effects of the human visual system. Such effects include the response to varying illumination, glare and differences between photopic and scotopic vision. This thesis evaluates several different tone mapping methods to allow a greater dynamic range to be used in real time visualisations. Several tone mapping methods have been implemented in the Avalanche Game Engine and evaluated using a small test group. To increase immersion in the visualization several filters aimed to simulate perceptual effects has also been implemented. The primary goal of these filters is to simulate scotopic vision. The tests showed that two tone mapping methods would be suitable for the environment used in the tests. The S-curve tone mapping method gave the best result while the Mean Value method gave good results while being the simplest to implement and the cheapest. The test subjects agreed that the simulation of scotopic vision enhanced the immersion in a visualization. The primary difficulties in this work has been lack of dynamic range in the input images and the challenges in coding real time graphics using a graphics processing unit.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-06-03T12:39:45Z No. of bitstreams: 1 claudiorobertobarbosasimoesrodrigues.pdf: 5762936 bytes, checksum: 8a7ae73391a9d56af3c18605a924763d (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-07-02T13:21:41Z (GMT) No. of bitstreams: 1 claudiorobertobarbosasimoesrodrigues.pdf: 5762936 bytes, checksum: 8a7ae73391a9d56af3c18605a924763d (MD5) / Made available in DSpace on 2016-07-02T13:21:41Z (GMT). No. of bitstreams: 1 claudiorobertobarbosasimoesrodrigues.pdf: 5762936 bytes, checksum: 8a7ae73391a9d56af3c18605a924763d (MD5) Previous issue date: 2012-12-17 / No intuito de aportar contribuições ao uso de diodos emissores de luz (ou LEDs, do inglês Light Emmiting Diodes) em sistemas de iluminação pública, cobrindo lacunas existentes no cenário nacional, este trabalho traz uma ampla análise sobre este tema. São apresentados conceitos básicos da fotometria clássica e um estudo sobre adaptações da fotometria para baixos níveis de luminância, que é a condição observada nas vias públicas durante o período noturno. Propõe-se a utilização de métricas escotópicas em sistemas de iluminação pública, uma vez que estas melhor representam a real sensação visual (se comparadas às métricas clássicas – fotópicas) nas condições de iluminamento típicas observadas nestes sistemas. Aspectos gerais da iluminação pública e um histórico da evolução do sistema nacional são brevemente apresentados bem como uma discussão sobre a situação atual deste sistema e das tecnologias utilizadas. O trabalho traz também um estudo sobre características ópticas, elétricas e mecânicas dos LEDs e uma análise dos diferentes tipos de LEDs brancos mais utilizados (HB – High Brightness – e HP – High Power – LEDs). Algumas normas e recomendações aplicáveis à utilização de LEDs em iluminação pública são analisadas e discutidas, com o intuito de se investigar lacunas ainda não cobertas pela normatização existente. Análises de resultados de ensaios de laboratório e de campo, inéditos no país, realizados com alguns modelos de luminárias LED para iluminação pública disponíveis no mercado, são também apresentadas. Um estudo sobre acionamento de LEDs é realizado, abordando técnicas de equalização de corrente entre arranjos série de LEDs associados em paralelo e diferentes tipos de circuitos de acionamento para LEDs. São propostas, analisadas e implementadas uma estrutura inédita de baixo custo para a equalização de corrente e um circuito para o acionamento de LEDs baseado na integração de dois conversores cc cc não isolados. A topologia de dois estágios proposta é composta por um conversor boost no primeiro estágio, utilizado para a correção do fator de potência e por um conversor buck no segundo estágio, utilizado para o controle da corrente nos LEDs. Por fim, o acompanhamento da implantação e do desempenho, ao longo de dez meses, do projeto piloto de iluminação pública empregando LEDs instalado no anel viário da Faculdade de Engenharia da UFJF é apresentado. / In order to present contributions to the use of light emitting diodes (LEDs) in public lighting systems, covering gaps on the national scenario, this work encompasses abroad analysis related to this issue. Basic concepts of classical photometry are presented and so a study of photometry’s adaptations for low luminance levels. It is proposed the use of scotopic metrics for street lighting systems, since they can better represent the actual visual sensation (when compared to the classical one – photopic metrics) on its typical nighty conditions. General aspects of street lighting and the history of the national’s public lighting system are briefly presented as well as a discussion about the current situation of the Brazilian system and technologies adopted. The work also brings a study of mechanical, electrical and optical characteristics of LEDs and an analysis of the two different types of commonly used white LEDs (HB – High Brightness – and HP – High Power - LEDs). Some standards and recommendations related to the use of LEDs in street lighting are analyzed and criticized, with the aim of show some gaps not yet covered by the existent standards. An unprecedented analysis regarding laboratory and field test benchmarking, taken for some luminaires available in commerce, is also presented. A study of LED driving is presented, covering equalization techniques for parallel LEDs strings and different types of LED driving circuits. It was proposed, analyzed and implemented a novel and low cost structure for current equalization and an LED driver, based on the integration of two non-isolated dc dc converters. This proposed two stages topology is composed by a boost converter, on the first stage, employed to the power factor correction and a buck converter, on the second stage, to control the LED’s current. Finally, the process of monitoring the installation and the performance, during ten months, of the LEDs public lighting pilot project, on the roadway belt of Engineering Faculty of UFJF is presented.