Arbetet undersöker grafikkortets lämplighet att tolka L-system i realtid. L-system är ett strängomskrivningssystem som beskriver strukturen hos botaniska former. Lsystem tolkas med hjälp av sköldpaddstolkningen för att få en geometrisk tolkning. För att utvärdera om grafikkortet är lämpat att tolka L-system i realtid har två system implementerats. Det ena systemet tolkar L-system på processorn och det andra systemet tolkar L-system på grafikkortet. Jämförelser har gjorts mellan dessa två system, framförallt av tidsåtgången för att tolka och rendera L-system. Men även minnesmängden som behövs för att realisera båda systemen i olika fall. Resultatet är en algoritm på grafikkortet som kan tolka L-system i realtid under rätt förutsättningar, men slutsatsen är att processorn är mer lämpad för uppgiften. Algoritmen på grafikkortet presterar väl vid hög polygonnivå, men under dessa förutsättningar utförs inte tolkningen i realtid. Vidare minskar algoritmen minnesåtgången på grafikkortet jämfört med algoritmen på processorn.

<p>Arbetet undersöker grafikkortets lämplighet att tolka L-system i realtid. L-system är ett strängomskrivningssystem som beskriver strukturen hos botaniska former. Lsystem tolkas med hjälp av sköldpaddstolkningen för att få en geometrisk tolkning. För att utvärdera om grafikkortet är lämpat att tolka L-system i realtid har två system implementerats. Det ena systemet tolkar L-system på processorn och det andra systemet tolkar L-system på grafikkortet. Jämförelser har gjorts mellan dessa två system, framförallt av tidsåtgången för att tolka och rendera L-system. Men även minnesmängden som behövs för att realisera båda systemen i olika fall. Resultatet är en algoritm på grafikkortet som kan tolka L-system i realtid under rätt förutsättningar, men slutsatsen är att processorn är mer lämpad för uppgiften. Algoritmen på grafikkortet presterar väl vid hög polygonnivå, men under dessa förutsättningar utförs inte tolkningen i realtid. Vidare minskar algoritmen minnesåtgången på grafikkortet jämfört med algoritmen på processorn.</p>

As imagens de angiografia por ressonância magnética (angio-RM) ou por tomografia computadorizada (angio-TC) permitem uma análise minuciosa das redes vasculares. A segmentação de redes vasculares a partir de tais imagens é uma das tarefas iniciais no diagnóstico de doenças vasculares como estenoses ou aneurismas. Porém, a grande diversidade de arquiteturas dos vasos dificulta a validação dos algoritmos de segmentação. Assim, a construção de redes vasculares sintéticas realistas permitem validar novas metodologias de segmentação de vasos. Este trabalho descreve uma metodologia de geração de redes vasculares sintéticas em três dimensões utilizando sistemas de Lindenmayer (L-systems) estocásticos. Para atingir esse objetivo, foram implementados um analisador léxico, um analisador sintático e um gerador de L-systems para a criação de vasos sintéticos baseado em gramáticas. A parametrização destas gramáticas possibilita a simulação de características naturais de vasos reais como o ângulo de bifurcação, comprimento, diâmetro médio e possibilita a simulação de anomalias vasculares. As expressões resultantes são utilizadas para criar imagens angiográficas sintéticas que simulam a distribuição de intensidades dos vasos em imagens angio-RM e angio-TC reais. As redes vasculares sintéticas podem também ser delimitadas por superfícies 3D arbitrárias de forma similar à geometria de órgãos. A flexibilidade de parametrização e natureza estocástica desta metodologia faz com que ela se torne uma ferramenta ideal para a validação de algoritmos de segmentação de vasos em imagens angiográficas. / Magnetic resonance angiography (MRA) or computed tomography angiography (CTA) images allow for a thorough analysis of the blood vessels. Vessel segmentation from MRA or CTA is thus the primary task in the diagnosis of vascular diseases such as stenosis and aneurysms. The wide architectural variability of the blood vessels, however, hinders the validation of vascular segmentation methods. The construction of synthetic realistic vascular architecture trees will aid in the validation of new vessel segmentation methodologies. This thesis describes a three-dimensional synthetic blood vessel generation methodology that employs stochastic Lindenmayer systems (L-systems). For this purpose, we implemented a parser and a generator of L-systems to create grammars that represent blood vessel architectures. The parameterization of the grammar allows one to simulate natural features of real vessels such as bifurcation angle, average length and diameter, and also accounts for vascular anomalies. The resulting expressions are used to create synthetic angiographic images that mimic real vessel intensity distributions in MRA and CTA. Blood vessel growth can also be delimited by arbitrary 3D surfaces that may represent organ geometries. The flexibility in the parameterization and stochastic nature of this methodology makes it an ideal tool for the validation of blood vessel segmentation algorithms from angiographic images.

As imagens de angiografia por ressonância magnética (angio-RM) ou por tomografia computadorizada (angio-TC) permitem uma análise minuciosa das redes vasculares. A segmentação de redes vasculares a partir de tais imagens é uma das tarefas iniciais no diagnóstico de doenças vasculares como estenoses ou aneurismas. Porém, a grande diversidade de arquiteturas dos vasos dificulta a validação dos algoritmos de segmentação. Assim, a construção de redes vasculares sintéticas realistas permitem validar novas metodologias de segmentação de vasos. Este trabalho descreve uma metodologia de geração de redes vasculares sintéticas em três dimensões utilizando sistemas de Lindenmayer (L-systems) estocásticos. Para atingir esse objetivo, foram implementados um analisador léxico, um analisador sintático e um gerador de L-systems para a criação de vasos sintéticos baseado em gramáticas. A parametrização destas gramáticas possibilita a simulação de características naturais de vasos reais como o ângulo de bifurcação, comprimento, diâmetro médio e possibilita a simulação de anomalias vasculares. As expressões resultantes são utilizadas para criar imagens angiográficas sintéticas que simulam a distribuição de intensidades dos vasos em imagens angio-RM e angio-TC reais. As redes vasculares sintéticas podem também ser delimitadas por superfícies 3D arbitrárias de forma similar à geometria de órgãos. A flexibilidade de parametrização e natureza estocástica desta metodologia faz com que ela se torne uma ferramenta ideal para a validação de algoritmos de segmentação de vasos em imagens angiográficas. / Magnetic resonance angiography (MRA) or computed tomography angiography (CTA) images allow for a thorough analysis of the blood vessels. Vessel segmentation from MRA or CTA is thus the primary task in the diagnosis of vascular diseases such as stenosis and aneurysms. The wide architectural variability of the blood vessels, however, hinders the validation of vascular segmentation methods. The construction of synthetic realistic vascular architecture trees will aid in the validation of new vessel segmentation methodologies. This thesis describes a three-dimensional synthetic blood vessel generation methodology that employs stochastic Lindenmayer systems (L-systems). For this purpose, we implemented a parser and a generator of L-systems to create grammars that represent blood vessel architectures. The parameterization of the grammar allows one to simulate natural features of real vessels such as bifurcation angle, average length and diameter, and also accounts for vascular anomalies. The resulting expressions are used to create synthetic angiographic images that mimic real vessel intensity distributions in MRA and CTA. Blood vessel growth can also be delimited by arbitrary 3D surfaces that may represent organ geometries. The flexibility in the parameterization and stochastic nature of this methodology makes it an ideal tool for the validation of blood vessel segmentation algorithms from angiographic images.

Research on modeling trees and plants has attracted a great deal of attention in recent years. Early procedural tree modeling can be divided into four main categories: rule-based algorithms, repetitive patterns, cellular automata, and particle systems. These methods offer a very high level of realism; however, creating millions of varied tree datasets manually is not logistically possible, even for professional 3D modeling artists. Trees created using these previous methods are typically static and the controllability of these procedural tree models is low. Deep generative models are capable of generating any type of shape automatically, making it possible to create 3D models at large scale. In this paper, we introduce a novel deep generative model that generates 3D (botanical) tree models, which are not only edible, but also have diverse shapes. Our proposed network, denoted BranchNet, trains the tree branch structures on a hierarchical Variational Autoencoder (VAE) that learns new generative model structures. By directly encoding shapes into a hierarchy graph, BranchNet can generate diverse, novel, and realistic tree structures. To assist the creation of tree models, we create a domain-specific language with a GUI for modeling 3D shape structures, in which the continuous parameters can be manually edited in order to produce new tree shapes. The trees are interpretable and the GUI can be edited to capture the subset of shape variability.

本篇論文對於傳統的東方風格雲紋進行分析，歸納其中的生成規則與排列特性，並將雲紋圖樣作簡單的分類，發現雲紋圖樣具有自我相似與遞迴組成的特性，此點特性非常符合原本用於模擬植物生長的L-system，所以我們就針對所蒐集的眾多雲紋圖樣參考資料所分析出的資訊定義了一個適用於立體東方風格雲紋的生成生長規則，以L-system之技術為基礎進行撰寫規則語法，利用所定義結構化之規則循序地在三維空間中動態模擬雲紋的生長與變化。更進一步地，加入具時間序列的參數調整，可以使得雲紋隨著時間軸的變化產生東方風格立體動態雲紋。 / Cloud Pattern is an important and common element in oriental decoration art, it is a challenge to extend the 2D abstract line into 3D objects. In this thesis, we try to induce the pattern generation rule and the feature of oriental cloud by analyzing the ancient oriental cloud pattern. We make classification from the oriental 2D cloud pattern, and discover the pattern is combined with symmetric, self-similar and repetitive features. These characteristic are the core of L-System, which is a suitable framework for plant growth and simulation, so we generate a new rule with specific parameter from the feature of 2D cloud pattern to generate oriental cloud pattern in 3-dimention space, simulating the growth of cloud and variety in structure. Further, we add the time temporal parameter to control the 3D oriental cloud pattern generation process dynamically.

Context. Procedurally generated content is a common way for companiesto save time and resources when creating modern video games.However, with generation of content there is often many variables controllingthe outcome that is tricky to get right. Objectives. The aim of this thesis is to research how people perceivethe branch density of trees used in video games. This will be done bycreating a foliage generation tool that is based on the functionality ofan L-system. The tool will be used to generate 12 trees with differentbranch densities which will be used for the questionnaire in the userexperiment. Method. A foliage generation tool has been created to be able togenerate trees with different branch and twig densities. The trees generatedwith the tool have been used in a user experiment in order tocome to a conclusion on how people perceive different branch densityin virtual trees. Result. Trees with higher branch and twig densities were rated higherthan trees with lower ones. The best rated tree in all regards was atree in the group with the second most main branches and had 0.6 intwig density. Conclusions. The results show that trees with a higher branch densitymade participants rate it as better looking in the user experiment.Although, trees with a high number of main branches and low twigdensity got the lowest score.

<p>Rapporten undersöker om kontextberoende generering av träd är lämpligt att använda i dataspel. I arbetet presenteras ett system som möjliggör procedurell generering av träd som omgivningen kan påverka. De faktorer som inverkar på trädets grenstruktur är solljus, vind och gravitation. Projektet bidrar med en lösning till att minska utvecklingstiderna för dataspel genom att möjliggöra att trädmodeller algoritmiskt genereras i datorn istället för att de modelleras av en artist.</p>

Accurate placement of pesticide droplets on to crop and weed surfaces is a key step in controlling pests and weeds in agricultural production systems. Droplets impact on plant surfaces, depositing pesticides that give protective coverage on crops and destructive coverage for insect or fungi pests and weeds. Coverage is complex and is determined by a multitude of interactions between factors such as size and density of spray droplets, relative humidity and turbulence of the air through which the droplets travel, and the physical characteristics of the target plant leaves, branches and stems that go to make up the architecture of their canopy. There are, however, concerns over the effect of pesticides on the environment and public health. This thesis combines three dimensional (3-D) computer modelling techniques, physical measurements of droplet movement and impact on a canopy in a wind tunnel and risk management techniques to maximise the effectiveness of pesticides and enable risks to public health and the environment from agricultural spraying activities to be minimised. L-studio, a Windows based software environment for creating simulation models of plant architecture was used in this study. A particle trajectory model, based on the combined ballistic and random-walk approach proposed by Mokeba et al. (1997) and Cox et al. (2000) was used to model spray droplet movement. Algorithms were included in the spray model to account for evaporation of droplets, entrained air and movement of air around the spray, collection efficiency, and droplet splash. Existing functional-structural plant models of cotton (Gossypium hirsutum L.) and sow thistle (Sonchus oleraceus L.) and a static empirical model of immature grass weeds have been combined with the spray model. An environmental program has been used to take the location of the leaves in 3-D space from the plant model and determine if spray droplets will impact on them. Wind tunnel measurements were made to determine initial droplet properties (droplet size, velocity, trajectory, density and fluid properties) and droplet impact characteristics (retention and splash). The results from these measurements were then used to define parameters within the spray model. Additional experiments to measure spray drift and spray deposition on various plant surfaces within the wind tunnel were used to evaluate the combined spray and plant architecture model. The combined spray and plant architectural model developed and evaluated in this thesis has provided a new method to study the influence of plant architecture on spray distribution. This work has shown that 3-D plant architecture can influence the amount of spray depositing on leaf surfaces. Deposition on plant surfaces was also found to increase with decreased wind speed and reduced release height. Droplet size did not have a significant influence on spray deposition onto broadleaf plants such as cotton or sow thistle. There was however, a tendency for fine sprays to give a higher deposit on small, narrow grass leaves. Spray drift was found to increase with decreasing droplet size, increasing the range of droplet sizes generated by a nozzle, decreasing sheet velocity (initial velocity of droplets), increasing wind speed, increasing (difference between dry and wet bulb temperature), decreasing liquid density and increasing release height. The combined spray and architectural model has also enabled the study of how effects such as droplet splash and retention can influence the distribution of the spray. In this study it was found that there was little difference in modelled spray drift, amount of spray on the ground or amount of spray on the cotton plant between situations where most droplets splash on impact and where no droplet splash on impact. Although the total amount of spray retained on the plant surface was similar for both situations, it was found that there were more droplets on the plant under the splash scenario leading to better coverage of the spray over the plant. These results indicate that the majority of the smaller splash droplets were re-distributed onto other parts of the plant rather than becoming lost as spray drift or ground deposit. Decision trees have been used in this thesis to characterise the benefits and risks from pesticide applications. Results from model simulations and/or physical measurements are used to estimate the relative proportion of the spray depositing on plant surfaces, depositing on the ground and drifting away from the treatment area. These deposition results give a measure of pesticide exposure that can be incorporated into the risk management framework to investigate the influence of various application scenarios. The applicability of this approach is shown by the example of endosulfan sprayed on cotton to control Helicoverpa spp. The decision tree model can be used to quickly compare different scenarios, such as deciding which application method should be used. It can be used to effectively aid spray decisions to maximise the effectiveness of pesticides and minimise risks to public health and the environment from agricultural spraying activities.

Creating content for games costs a lot of time and money. This thesis will use the Lindenmayer system for procedural content creation inside the game to answer the question, how can L-systems be exploited for gameplay mechanic purposes? The application implemented in Unity 3D is a form of tree L-system with player controlled recursion, independent branches and poisonous branches. The recursion can allow the player to grow strategically, independent branches can be used as seed and grow on its own if they break o , poisonous branches to force players to think when navigating.