National Repository of Grey Literature
Not a member yet
574846 research outputs found
Sort by
Method for designing 2D digital FIR filters by transforming 1D digital FIR filters designed by the optimal Parks-McClellan method
Two-dimensional FIR filters are an important tool in digital signal processing, especially in the field of image data analysis and processing. This work focuses on the design of filters with zero phase and the required spatial symmetry using various methods using the Parks–McClellan optimization method. The resulting filters were implemented and tested in the MATLAB environment, including circular and directional synthesized structures. The work will hopefully open up possibilities for further development, for example by implementing them in Python or using neural networks
Gamebook
This bachelor thesis focuses on the creation of an original gamebook—an interactive book that combines literary text with game elements. The reader is not a passive recipient of the narrative but actively chooses their path through the fictional world. The resulting text takes the form of a collection of interconnected short stories that blend elements of cosmic horror and magical realism. The stories are characterized by diverse linguistic approaches and a strong grounding in a rural setting. The thesis also includes collaboration with an illustrator and the graphic layout of the book
DomainRadar Extension for Malicious Domain Name Detection using Web Page Contents
Tato bakalářská práce pojednává o metodách detekce phishingových, škodlivých a benigních domén na základě analýzy textového obsahu jejich webových stránek. Hlavní důraz je kladen na využití TF-IDF (Term Frequency - Inverse Document Frequency), metody, která umožňuje určit význam slov v dokumentech na základě jejich frekvence v textu a inverzní frekvence v celém datovém korpusu. Studie ukazuje, že obsah webových stránek obsahuje užitečné textové prvky, které lze použít k automatické klasifikaci domén. Na základě těchto vlastností byly vyvinuty a natrénovány modely klasifikátorů, které dosáhly přesností téměř 90% oba. Použití TF-IDF v kombinaci s metodami strojového učení umožňuje efektivně identifikovat phishing a škodlivé zdroje a také je odlišit od bezpečných domén. Výsledky potvrzují vysoký přínos analýzy textu při řešení problémů kybernetické bezpečnosti a lze je využít k vytvoření automatizovaných systémů pro monitorování a ochranu uživatelů na internetu
Preventive maintenance for unconventional waterjet technology
This thesis deals with the issue of preventive maintenance in unconventional waterjet cutting technology, specifically in the environment of AWAC, spol. s.r.o. The aim of the thesis is to analyze the current state of maintenance of machinery using abrasive waterjet technology, to identify critical areas with high wear rates and to propose an optimized preventive maintenance plan taking into account technical, operational and economic aspects. The theoretical part of the thesis describes the history, principle and distribution of waterjet methods, as well as the technical components of the equipment and the influence of individual technological parameters on the quality of the cut surface. The practical part focuses on the assessment of specific machines (e.g. Flow MACH 4C, MicroStep AquaCut 3001.40WW) and high-pressure pumps in terms of the frequency of failures and the effectiveness of service interventions. In addition, a comparison of water quality in different regions of the Czech Republic, an analysis of spare parts suppliers and a technical and economic evaluation of the proposed solution are presented. The result is a newly developed preventive maintenance plan in electronic form, which aims to increase operational reliability and reduce unplanned downtime of machinery and equipment. The work contributes to the reduction of operating costs and is fully applicable to other similar facilities
GPU accelerated algorithms for the shortest path problem
Výpočet nejkratších cest z jednoho zdroje je základním problémem v teorii grafů, jehož cílem je nalézt nejkratší cesty z daného výchozího uzlu ke všem ostatním uzlům v grafu. U velkých hustých grafů může být výpočet nejkratších cest i pomocí optimálních sekvenčních algoritmů časově velmi náročný a trvat i několik minut. Výpočet lze významně urychlit s využitím paralelismu. Paralelizace algoritmu může přinést vysokou míru zrychlení výpočtu vůči sekvenčním variantám. Jedním z přístupů je paralelizace na úrovni grafické karty. Pro hledání nejkratších cest lze využít velké množství jader, kterými karta disponuje. Tato práce přínáší nový přístup k paralelizaci algoritmu Dijkstra na úrovni grafické karty včetně jeho implementace, porovnání jeho efektivity vůči sekvenční implementaci a dvěma známým efektivním paralelizacím na úrovni grafické karty s jiným přístupem k algoritmu. Na hustých grafech nový přístup ve většině případů úspěšně poráží ostatní implementace díky efektivnímu využití enormního počtu jader grafické karty. Tato práce zdůrazňuje silné a slabé stránky každého z přístupů spolu s jejich vyhodnocením mezi sebou
System for change detection from satellite data using artificial intelligence
Earth observation plays a crucial role in various fields such as urban development or environmental monitoring. Fast and accurate interpretation of satellite imagery is essential for real-time decision-making as well as for long-term planning. Traditional data processing methods often lead to delays, high costs, and low efficiency. These limitations pose a significant obstacle in situations where rapid response is critical. This thesis focuses on the design and implementation of a system for automatic change detection in satellite imagery using artificial intelligence and deep learning techniques. The emphasis is placed on improving computational efficiency, which is essential for processing large volumes of data in resource-constrained environments, such as directly onboard satellites. The outcome of the work is a set of three models, evaluated using standard performance metrics. The results show that the highest value of Intersection over Union (IoU) was achieved by the U-net model (0.7794). More compact architectures based on MobileNetV2 exhibited lower IoU scores (0.7090 and 0.7383, respectively), but maintained competitive performance with significantly reduced memory requirements. These findings demonstrate that lightweight models can be suitable for change detection tasks when properly optimized
The impact of metabolites of flavonoids (phenylacetic acids) on the iron-catalysed Fenton reaction
Charles University Faculty of Pharmacy in Hradec Králové Department of Pharmacognosy and Pharmaceutical Botany Candidate: Kristýna Podlesná Supervisor: Zuzana Lomozová, PharmDr., Ph.D. Title of Thesis: The impact of metabolites of flavonoids (phenylacetic acids) on the iron-catalysed Fenton reaction Keywords: iron, flavonoids, phenylacetic acids, Fenton reaction, antioxidant, pro-oxidant Iron is a biogenic trace element that plays an irreplaceable role in the human body. Due to its ability to switch between the oxidation states of Fe2+ and Fe3+, it acts as a biocatalyst in chemical reactions. The amount of iron in the body is tightly regulated, as both deficiency and excess can result in a wide range of pathological conditions, including anemia or oxidative tissue damage. Flavonoids form a large group of secondary metabolites in plants. They are a common part of the human diet, and their consumption is associated with many health benefits. In particular, they exhibit antioxidant activity. However, under certain circumstances, they can also exert the opposite effect as pro-oxidants. Flavonoids are extensively metabolised before entering the systemic circulation. They are transformed into small phenolic substances, including phenylacetic acids. In this thesis, the effect of eight phenylacetic acids..
Otherness in children's literature of the 21st century from the perspective of teachers and pupils
This diploma thesis, entitled "The Theme of Bullying in Children's Literature of the 21st Century from the Perspective of Teachers and Pupils", deals in depth with bullying as a phenomenon that permeates the entire society. In connection with bullying in children of younger school age, the thesis also focuses on the classroom climate, which can be seen as one of the possible factors leading to the prevention of bullying in schools. The diploma thesis sets several interconnected goals. First of all, it focuses on finding out, through content analysis, how the theme of bullying is depicted in children's literature published in the 21st century. It also tries to investigate whether and how teachers at the first level of primary school work with this topic, what methods they use to mediate it, and what specific literary titles have proven successful in practice. Data were obtained on the basis of a questionnaire survey, in which a total of 45 respondents (first level primary school teachers) participated. Last but not least, the work focuses on the selection of suitable books reflecting the issue of bullying, the creation of literary lessons according to the three-phase E-U-R learning model, their implementation in teaching and subsequent evaluation of their effectiveness. The research selected and..
Formative assessment and family involvement in the assessment process
The diploma thesis, which is divided into two parts, theoretical and empirical, deals with formative assessment, defines this concept and describes the methods that can be used in formative assessment. It also focuses on cooperation between school and family. It describes the forms of communication and what roles parents of pupils can represent. In the last chapter of the theoretical part, we encounter the most common forms of cooperation between school and family, specifically class meetings and tripartite meetings. The empirical part includes semi-structured interviews with six teachers from different primary schools, who answer nineteen questions divided into three areas: formative assessment, family and tripartite meeting. Using the interviews, the thesis tries to answer whether parents can be involved in the assessment of their children. The interviewed teachers agreed that parental participation in the assessment of their children is very important for them to achieve their predetermined goals