Fyzikálně přijatelné osvětlení v reálném čase je často dosaženo použitím aproximací. Současné metody často aproximují globální osvětlení v prostoru obrazu s využitím schopností moderních grafických karet. Dva techniky z této kategorie, screen-space ambient occlusion a screen-space directional occlusion jsou popsány detailněji v této práci. Screen-space directional occlusion je zobecněná verze screen-space ambient occlusion s podporou jednoho difúzního odrazu a závislostí na směrové informaci světla. Hlavním cílem projektu bylo experimentování s těmito metodami. Pro uniformní distribuci náhodných vzorek pro obě metody byla použita Halton sekvence. Pro potlačení šumu je použita bilaterální filtrace, která bere do úvahy geometrické vlastnosti scény. Metody jsou dál zrychleny použitím nižších rozlišení pro výpočet. Rekonstrukce výsledků do původní velikosti pro vytvoření konečného obrazu je realizována pomoci joint bilateral upsamplingu. Kromě metod globálního osvětlení byly v práci použity aj metody pro mapování stínů a HDR osvětlení.

The target of this thesis is sumarize the theory of metods used for composition image sequence into single high dynamic range image. Selected methods and algorithms are described in theoretical part and implemented in practical part of this thesis. It was used C++ language with using OpenCV library.

Cílem této práce je demonstrovat možnost renderování srsti přímo do existujících obrazů bez toho, aby bylo po uživateli požadováno překreslení všech pixelů nebo dodání kompletní 3D geometrie a osvětlení. Srst je přidána na povrch objektů pomocí extrakce jejich přibližného tvaru a světelných informací z obrazu a takto získaný objekt je poté přerenderován. Tento přístup je nový v tom, že vysokoúrovňové úpravy obrazu (jako např. přidání srsti), mohou úspěšně vést k vizuálně korektním výsledkům a to i přes omezení nepřesnou geometrií a světelnými podmínkami. Relativně velká množina technik použitých v této práci zahrnuje obrazy s velkým dynamickým rozsahem, metody extrakce 3D tvaru z obrazu, výsledky výzkumu vnímání tvaru a osvětlení a fotorealistické renderování. Hlavním cílem práce je potvrdit koncept popsaný výše. Hlavním implementačním jazykem bylo C++ s použitím knihoven wxWidgets, OpenGL a libTIFF. Renderování bylo realizováno v software 3Delight kompatibilním se standardem Renderman, za pomoci množiny shaderů implementovaných v nativním jazyce Rendermanu.

<p>Imaging and computer vision systems deployed in real-world environments face the challenge of accommodating a wide range of lighting conditions. However, the cost, the demand for high resolution, and the miniaturization of imaging devices impose physical constraints on sensor design, limiting both the dynamic range and effective aperture size of each pixel. Consequently, conventional CMOS sensors fail to deliver satisfactory capture in high dynamic range scenes or under photon-limited conditions, thereby impacting the performance of downstream vision tasks. In this thesis, we address two key problems: 1) exploring the utilization of spatial multiplexing, specifically spatially varying exposure tiling, to extend sensor dynamic range and optimize scene capture, and 2) developing techniques to enhance the robustness of object detection systems under photon-limited conditions.</p> <p><br></p> <p>In addition to challenges imposed by natural environments, real-world vision systems are susceptible to adversarial attacks in the form of artificially added digital content. Therefore, this thesis presents a comprehensive pipeline for constructing a robust and scalable system to counter such attacks.</p>

Derived from and inspired by the adaptive immune system of bacteria, CRISPR has gone from basic biology knowledge to a revolutionizing biotechnological tool, applicable in many research areas such as medicine, industry and agriculture. The full mechanism of CRISPR-Cas9 was first published in 2012 and various CRISPR-Cas systems have already passed the first stages of clinical trials as new gene therapies. The immense research has resulted in continuously growing knowledge of CRISPR systems and the technique seems to have the potential to greatly impact all life on our planet. Therefore, this literature study aims to thoroughly describe the CRISPR-Cas system, and further suggest an undergraduate laboratory exercise involving gene editing with the CRISPR-Cas9 tool. In this paper, we describe the fundamental technical background of the CRISPR-Cas system, especially emphasizing the most studied CRISPR-Cas9 system, its development and applications areas, as well as highlighting its current limitations and ethical concerns. The history of genetic engineering and the discovery of the CRISPR system is also described, along with a comparison with other established gene editing techniques.  This study concludes that a deeper knowledge about CRISPR is important and required since the technique is applicable in many research areas. A laboratory exercise will not only inspire but also provide extended theoretical and practical knowledge for undergraduate students.

Die Modifikation von Genen ist in den molekularen Biowissenschaften ein fundamentales Werkzeug, um die Funktion von Genen zu studieren (Reverse Genetik). Diese Arbeit hat erfolgreich Zinkfinger- und CRISPR/Cas9-Nukleasen für die Verwendung in C. reinhardtii etabliert, um Gene im Kerngenom gezielt auszuschalten und präzise zu verändern. Basierend auf vorausgegangener Arbeit mit Zinkfingernukleasen (ZFN) konnte die Transformationseffizienz um das 300-fache verbessert werden, was die Inaktivierung von Genen auch in motilen Wildtyp-Zellen ermöglichte. Damit war es möglich, die Gene für das Kanalrhodopsin-1 (ChR1), Kanalrhodopsin-2 (ChR2) und das Chlamyopsin-1/2-Gen (COP1/2) einzeln und gemeinsam auszuschalten. Eine Analyse der Phototaxis in diesen Stämmen ergab, dass die Phototaxis durch Inaktivierung von ChR1 stärker beeinträchtigt ist als durch Inaktivierung von ChR2. Um das CRISPR/Cas9-System zu verwenden, wurden die Transformationsbedingungen so angepasst und optimiert, dass der Cas9-gRNA-Komplex als in vitro hergestelltes Ribonukleoprotein in die Zellen transformiert wurde. Um die Bedingungen für präzise Genmodifikationen zu messen und zu verbessern, wurde das SNRK2.2-Gen als Reportergen für eine „Blau-Grün Test“ etabliert. Kleine Insertionen von bis zu 30 bp konnten mit kurzen Oligonukleotiden eingefügt werden, während größere Reportergene (mVenus, SNAP-Tag) mithilfe eines Donor-Plasmids generiert wurden. In dieser Arbeit konnten mehr als 20 nicht-selektierbare Gene – darunter 10 der 15 potenziellen Photorezeptorgene – mit einer durchschnittlichen Mutationsrate von 12,1 % inaktiviert werden. Insgesamt zeigt diese Arbeit in umfassender Weise, wie Gen-Inaktivierungen und Modifikationen mithilfe von ZFNs und des CRISPR/Cas9-Systems in der Grünalge C. reinhardtii durchgeführt werden können. Außerdem bietet die Sammlung der zehn Photorezeptor-Knockouts eine aussichtsreiche Grundlage, um die Vielfalt der Photorezeptoren in C. reinhardtii zu erforschen. / Gene editing is a fundamental tool in molecular biosciences in order to study the function of genes (reverse genetics). This study established zinc-finger and CRISPR/Cas9 nucleases for gene editing to target and inactivate the photoreceptor genes in C. reinhardtii. In continuation of previous work with designer zinc-finger nucleases (ZFN), the transformation efficiency could be improved 300-fold, which enabled the inactivation of genes in motile wild type cells. This made it possible to disrupt the Channelrhodopsin-1 (ChR1), Channelrhodopsin-2 (ChR2) and Chlamyopsin-1/2 (COP1/2) genes individually and in parallel. Phototaxis experiments in these strains revealed that the inactivation of ChR1 had a greater effect on phototaxis than the inactivation of ChR2. To apply the CRISPR/Cas9 system, the transformation conditions were adapted and optimized so that the Cas9-gRNA complex was successfully electroporated into the cells as an in vitro synthesized ribonucleoprotein. This approach enabled gene inactivations with CRISPR/Cas9 in C. reinhardtii. In order to measure and improve the conditions for precise gene modifications, the SNRK2.2 gene was established as a reporter gene for a ‘Blue-Green test’. Small insertions of up to 30 bp were inserted using short oligonucleotides, while larger reporter genes (mVenus, SNAP-tag) were integrated using donor plasmids. Throughout this study, more than 20 non-selectable genes were disrupted, including 10 of the photoreceptor genes, with an average mutation rate of 12,1 %. Overall, this work shows in a comprehensive way how gene inactivations and modifications can be performed in green alga C. reinhardtii using ZFNs or CRISPR/Cas9. In addition, the collection of the ten photoreceptor knockouts provides a promising source to investigate the diversity of photoreceptor genes in C. reinhardtii.

The physics of ultracold quantum gases has been the subject of a long-lasting and intense research activity, which started almost a century ago with purely theoretical studies and had a fluorishing experimental development after the implementation of laser and evaporative cooling techniques that led to the first realization of a Bose Einstein condensate (BEC) over 25 years ago. In recent years, a great interest in ultracold atoms has developed for their use as platforms for quantum technologies, given the high degree of control and tunability offered by ultracold atom systems. These features make ultracold atoms an ideal test bench for simulating and studying experimentally, in a controlled environment, physical phenomena analogous to those occurring in other, more complicated, or even inaccessible systems, which is the idea at the heart of quantum simulation. In the rapidly developing field of quantum technologies, it is highly important to acquire an in-depth understanding of the state of the quantum many-body system that is used, and of the processes needed to reach the desired state. The preparation of the system in a given target state often involves the crossing of second order phase transitions, bringing the system strongly out-of-equilibrium. A better understanding of the out-of-equilibrium processes occurring in the vicinity of the transition, and of the relaxation dynamics towards the final equilibrium condition, is crucial in order to produce well-controlled quantum states in an efficient way. In this thesis I present the results of the research activity that I performed during my PhD at the BEC1 laboratory of the BEC center, working on ultracold gases of 23Na atoms in an elongated harmonic trap. This work had two main goals: the accurate determination of the equilibrium properties of a Bose gas at finite temperature, by the measurement of its equation of state, and the investigation of the out-of-equilibrium dynamics occurring when a Bose Einstein condensate is prepared by cooling a thermal cloud at a finite rate across the BEC phase transition.To study the equilibrium physics of a trapped atomic cloud, it is crucial to be able to observe its density distribution in situ. This requires a high optical resolution to accurately obtain the density profile of the atomic distribution, from which thermodynamic quantities can then be extracted. In particular, in a partially condensed atomic cloud at finite temperature, it is challenging to resolve well also the boundaries of the BEC, where the condensate fraction rapidly drops in a narrow spatial region. This required an upgrade of the experimental apparatus in order to obtain a high enough resolution. I designed, tested and implemented in the experimental setup new imaging systems for all main directions of view. Particular attention was paid for the vertical imaging system, which was designed to image the condensates in trap with a resolution below 2 μm, with about a factor 4 improvement compared to the previous setup. The implementation of the new imaging systems involved a partial rebuilding of the experimental apparatus used for cooling the atoms. This created the occasion for an optimization of the whole system to obtain more stable working conditions. Concurrently I also realized and included in the experiment an optical setup for the use of a Digital Micromirror Device (DMD) to project time-dependent arbitrary light patterns on the atoms, creating optical potentials that can be controlled at will. The use of this device opens up exciting future scenarios where it will be possible to locally modify the trapping potential and to create well-controlled barriers moving through the atomic cloud. Another challenge in imaging the density distribution in situ is determined by the fact that the maximum optical density (OD) of the BEC, in the trap center, exceeds the low OD of the thermal tails by several orders of magnitude. In order to obtain an accurate image of the whole density profile, we developed a minimally destructive, multi-shot imaging technique, based on the partial transfer of a fraction of atoms to an auxiliary state, which is then probed. Taking multiple images at different extraction fractions, we are able to reconstruct the whole density profile of the atomic cloud avoiding saturation and maintaining a good signal to noise ratio. This technique, together with the improvements in the imaging resolution, has allowed us to accurately obtain the optical density profile of the Bose gas in trap, from which the 3D density profile was then calculated applying an inverse Abel transform, taking advantage of the symmetry of the trap. From images of the same cloud after a time-of-flight expansion, we measured the temperature of the gas. From these quantities we could find the pressure as a function of the density and temperature, determining the canonical equation of state of the weakly interacting Bose gas in equilibrium at finite temperature. These measurements also allowed us to clearly observe the non-monotonic temperature behavior of the chemical potential near the critical point for the phase transition, a feature that characterizes also other superfluid systems, but that had never been observed before in weakly interacting Bose gases. The second part of this thesis work is devoted to the study of the dynamical processes that occur during the formation of the BEC order parameter within a thermal cloud. The cooling at finite rate across the Bose-Einstein condensation transition brings the system in a strongly out-of-equilibrium state, which is worth investigating, together with the subsequent relaxation towards an equilibrium state. This is of interest also in view of achieving a better understanding of second order phase transitions in general, since such phenomena are ubiquitous in nature and relevant also in other platforms for quantum technologies. A milestone result in the study of second order phase transitions is given by the Kibble-Zurek mechanism, which provides a simple model capturing important aspects of the evolution of a system that crosses a second-order phase transition at finite rate. It is based on the principle that in an extended system the symmetry breaking associated with a continuous phase transition can take place only locally. This causes the formation of causally disconnected domains of the order parameter, at the boundaries of which topological defects can form, whose number and size scale with the rate at which the transition is crossed, following a universal power law. It was originally developed in the context of cosmology, but was later successfully tested in a variety of systems, including superfluid helium, superconductors, trapped ions and ultracold atoms. The BEC phase transition represents in this context a paradigmatic test-bench, given the high degree of control at which this second-order phase transition can be crossed by means of cooling ramps at different rates. Already early experiments investigated the formation of the BEC order parameter within a thermal cloud, after quasi-instantaneous temperature quenches or very slow evaporative cooling. In the framework of directly testing the Kibble-Zurek mechanism, further experiments were performed, both in 2D and 3D systems, focusing on the emergence of coherence and on the statistics of the spontaneously generated topological defects as a function of the cooling rate. The Kibble-Zurek mechanism, however, does not fully describe the out-of-equilibrium dynamics of the system at the transition, nor the post-quench interaction mechanisms between domains that lead to coarse-graining. Most theoretical models are based on a direct linear variation of a single control parameter, e.g. the temperature, across the transition. In real experiments, the cooling process is controlled by the tuning of other experimental parameters and a global temperature might not even be well defined, in a thermodynamic sense, during the whole process. Moreover, the temperature variation is usually accompanied by the variation of other quantities, such as the number of atoms and the collisional rate, making it difficult to accurately describe the system and predict the post-quench properties. Recent works included effects going beyond the Kibble-Zurek mechanism, such as the inhomogeneity introduced by the trapping potential, the role of atom number losses, and the saturation of the number of defects for high cooling rates. These works motivate further studies, in particular of the dynamics taking place at early times, close to the crossing of the critical point. The aim of the work presented in this thesis is to further investigate the timescales associated to the formation and evolution of the BEC order parameter and its spatial fluctuations, as a function of the rate at which the transition point is crossed. We performed experiments producing BECs by means of cooling protocols that are commonly used in cold-atom laboratories, involving evaporative cooling in a magnetic trap. We explored a wide range of cooling rates across the transition and found a universal scaling for the growth of the BEC order parameter with the cooling rate and a finite delay in its formation. The latter was already observed in earlier works, but for a much more limited range of cooling rates. The evolution of the fluctuations of the order parameter was also investigated, with an analysis of the timescale of their decay during the relaxation of the system, from an initial strongly out-of-equilibrium condition to a final equilibrium state. This thesis is structured as follows: The first chapter presents the theoretical background, starting with a brief introduction to the concept of Bose Einstein condensation and a presentation of different models describing the thermodynamics of an equilibrium Bose gas. The second part of this chapter then deals with the out-of-equilibrium dynamics that is inevitably involved in the crossing of a second-order phase transition such as the one for Bose-Einstein condensation. The Kibble-Zurek mechanism is briefly reviewed and beyond KZ effects are pointed out, motivating a more detailed investigation of the timescales involved in the BEC formation. In the second chapter, I describe the experimental apparatus that we use to cool and confine the atoms. Particular detail is dedicated to the parts that have been upgraded during my PhD, such as the imaging system. In the third chapter I show our experimental results on the measurement of the equation of state of the weakly interacting uniform Bose gas at finite temperature. In the fourth chapter I present our results on the out-of-equilibrium dynamics in the formation of the condensate order parameter and its spatial fluctuations, as a function of different cooling rates.