Institut für Computervisualistik
Refine
Year of publication
Document Type
- Bachelor Thesis (100) (remove)
Keywords
- Computervisualistik (8)
- OpenGL (7)
- Computergrafik (5)
- Android (4)
- Graphik (4)
- Line Space (4)
- Augmented Reality (3)
- C++ (3)
- Compute Shader (3)
- Computergraphik (3)
Colonoscopy is the gold standard for the detection of colorectal polyps that can progress into cancer. In such an examination, physicians search for polyps in endoscopic images. Thereby polyps can be removed. To support experts with a computer-aided diagnosis system, the University of Koblenz-Landau currently makes some efforts in research different methods for automatic detection. Comparable to traditional pattern recognition systems, features are initially extracted and a classifier is trained on such data. Afterwards, unknown endoscopic images can be classified with the previously trained classifier. This thesis concentrates on the extension of the feature extraction module in the existing system. New detection methods are compared to existing techniques. Several features are implemented, incorporating Graylevel Co-occurrence Matrices, Local Binary Patterns and Discrte Wavelet Transform. Different modifications on those features are applied and evaaluated.
This thesis deals with the development of an authoring system for modeling 3D environments with physical description. In contrast to creating scenes in other common modeling tools, one can now compute and describe physical entities of a scene additional to the usual geometry. It is very important for those authoring systems to be extendable and customizable for specific requirement of the user. The focus lies on developing simple program architecture, which is easy to extend and to modify.
There are a few systems high and low-cost ones for gaze tracking. Normally low-cost systems go in hand with low-resolution cameras. Here the image quality is poor, so the algorithms for detecting the gaze have to work more precisely. But how to test and analyse them correctly, when there is a bad image quality and no reference point known? The idea of this work is, to generate synthetic eye images, where the reference points are known, because they are mainly manually set and then to test and analyse the algorithms with these synthetic images. By switching on features like gaussian noise or a second glint-like reflection point, it is possible to stepwise approximate the synthetic images close to reality. In fact the experiments will lead to an improvement of the algorithms used in a low-resolution system environment.
The present work starts with an introduction of methods for three-dimensional curve skeletonization. Different kinds of historic and recent skeletonization approaches are analysed in detail. Later on, a state-of-the-art skeletonization algorithm is introduced. This algorithm deals as a basis for the own approach presented subsequently. After the description and definition of a new method improving the state-of-the-art algorithm, experiments are conducted to get appraisable results. Next, a ground truth is described which has been set up manually by humans. The human similarity evaluations are compared with the results of the automatic computer-based similarity measures provided by the own approach. For this comparison, standard evaluation criteria from the field of information retrieval have been used.
Texture-based text detection in digital images using wavelet features and support vector machines
(2010)
In this bachelor thesis a new texture-based approach for the detection of text in digital images is presented. The procedure can be essentially divided into two main tasks, in detection of text blocks and detection of individual words, whereby the individual words are extracted from the detected text blocks. Roughly, the developed method acts with multiple support vector machines, which classify possible text regions of an image into real text regions, using wavelet-based features. In the process the possible text regions are defifined by edge projections with diσerent orientations. The results of the approach are X/Y coordinates, width and height of rectangular regions of an image, which contains individual words. This knowledge can be further processed, for example by an optical character recognition software to get the important and useful text information.
The following thesis analyses the functionality and programming capabilitiesrnof compute shaders. For this purpose, chapter 2 gives an introductionrnto compute shaders by showing how they work and how they can be programmed. In addition, the interaction of compute shaders and OpenGL 4.3 is shown through two introductory examples. Chapter 3 describes an NBodyrnsimulation that has been implemented in order to show the computational power of compute shaders and the use of shared memory. Then it is shown in chapter 4 how compute shaders can be used for physical simulationsrnand where problems may arise. In chapter 5 a specially conceived and implemented algorithm for detecting lines in images is described and then compared with the Hough transform. Lastly, a final conclusion is drawn in chapter 6.
Das Kommunikationsverhalten hat sich in den letzten Jahren durch die Smartphonenutzung verändert. Die Nutzer kommunizieren oft nur noch über den elektronischen Weg. Die persönliche Kommunikation, außerhalb des Smartphones, nimmt ab. Das Umfeld gerät unterdessen in Vergessenheit. In der vorliegenden Arbeit werden verschiedene Spielkonzepte entwickelt, welche die Kommunikation steigern sollen. Realisiert wird der Ansatz in einer prototypischen Stadtführer-App, nach den Spielkonzepten von "Scotland Yard" und "Schnitzeljagd". Während der Nutzung müssen die Spieler verschiedene Aufgaben lösen. Welches Spielkonzept sich in Bezug auf die Kommunikationsförderung am besten eignet, wird in einer Evaluation analysiert.
Der Markt der mobilen Endgeräte entwickelt sich schnell weiter und es kommen schon Kinder im frühsten Alter mit solchen Technologien in Berührung. Daher ist es wichtig, Kinder richtig an die Geräte heranzuführen. Von Vorteil wäre eine Einbindung von Smartphones und Tablets, im Bezug auf den Lernprozess, in den Unterricht. Die vorliegende Arbeit behandelt deshalb das Konzept einer Lernspielapp, die durch Pädagogen konfiguriert werden kann. Die Evaluation soll Aufschluss über die Motivation der Kinder geben und die Aufgeschlossenheit der Pädagogen gegenüber neuen Medien ermitteln.
3D-models are getting more important in many areas such as multimedia applications, robotics or film industries. Of particular interest is the creation of 3D-models from a series of monocular images. This is because the cameras that are required for this purpose are becoming cheaper, smaller and more sophisticated at the same time. Increasingly often, suitable cameras are already integrated in devices like smartphones, tablet PCs or cars for example. Hence, there is a great potential for applications of this reconstruction technique.
This thesis is based on the use of a series of images that were taken with arncalibrated camera. The first step is to extract point correspondences from this image series making use of the well-known SURF- and A-KAZE-features. Starting from the point correspondences, it is possible to reconstruct a 3D-Modell with different algorithms that consists of a point cloud and camera poses. To reduce errors in the 3D-model, this thesis especially focuses on explaining the bundle adjustment algorithm, which is being used for a non-linear error minimization of a cost function.
The thesis also introduces the application for the 3D-reconstruction and the visualization of the results, that was developed in the course of this thesis.
The implemented system is evaluated based on statistics and the newly aquiredrnknowledge is presented. The thesis concludes with a summary of its results, and a number of ideas for potential future applications and developments.
This thesis imparts a general view of the mechanics and implementation of latest voxelization strategies using the GPU. In addition to established voxelization procedures using the rasterization pipeline, new possibilities arising from GPGPU programming are examined. On the basis of the programming language C++ and the graphics library OpenGL the implementation of several methods is explained. The methods are compared in terms of performance and quality of the resulting voxelization and are evaluated critically with regards to possible use cases. Furthermore, two exemplary applications are detailed that use a voxelized scene in such a way that the augmentation of established techniques of real time graphics are facilitated. To this end, the concepts and the implementations of Transmittance Shadow Mapping and of Reflective Shadow Mapping utilizing a voxel based ambient occlusion effect is explained. Finally, the prolonging relevance of voxelization is put into prospect, by addressing latest research and further enhancements and applications of the presented methods.
Meine Bachelorarbeit ist als Fortsetzung und Weiterentwicklung der Studienarbeit "Erstellung einer HTML 5-basierten Lernoberflache zur Unterstützung der Zielgruppe der AnfängerInnen für das Softwareprogramm CrypTool 2" von Liane Syre, eingereicht am 13.08.2013 an der Universität Koblenz-Landau, konzipiert. Die Inhalte ihrer Studienarbeit dienen somit als Grundlage für die hier vorliegende Bachelorarbeit.
Ich habe im Rahmen meiner Bachelorarbeit einen Prototyp eines teilweise automatisierten Lerntagebuchs entwickelt, welchen ich im Folgenden Vorstellen mochte. Außerdem stelle ich meine ebenfalls eigenständig konzipierte LearnCrypTo-Homepage in dieser Arbeit vor.
Nach einer Einleitung in die Thematik (Kapitel 2) folgt eine Literaturstudie zu Serious Games als Teilbereich des Media-based Learning (Kapitel 3), in welchem auch explizit auf die Führung eines Lerntagebuches und didaktische Konzepte eingegangen wird.
Es folgt eine Beschreibung wichtiger Aspekte bei der Erstellung einer Lernsoftware und gehe im Kontext der Serious Games noch mal auf die Spielerstellung ein (Kapitel 4).
Da meine Bachelorarbeit vor allem darauf ausgerichtet ist, CrypTool 2 AnfängerInnen nahe zu bringen, erläutere ich sodann die für benannte Zielgruppe wichtigsten Funktionen von CrypTool 2 (Kapitel 5).
Dann gehe ich naher auf Inhalt und Aufbau meiner eigens erstellten LearnCrypTo-Homepage ein (Kapitel 6).
Ich erläutere gegen Schluss meiner Arbeit ausgewählte technische Aspekte (Kapitel 7).
In einem abschließenden Fazit versuche ich, die Ergebnisse meiner Arbeit zu bewerten und wage einen kleinen Ausblick (Kapitel 8).
Im Verlaufe der Arbeit wird vermehrt mit einer hauptsachlich englischen Terminologie gearbeitet. Um Missverständnissen vorzubeugen werde ich die englischen Begriffe bei ihrer ersten Verwendung in meiner Arbeit ins Deutsche übersetzen und im Folgenden dann weiter mit der englischen Bezeichnung arbeiten.
Um Lesbarkeit zu fordern benutze ich in meiner Arbeit das generische Maskulinum.
Online Handschrifterkennung chinesischer Schriftzeichen auf androidfähigen mobilen Endgeräten
(2014)
Usage of mobile dictionaries or translators requires an input. This input has to be processed and recognized beforehand. Chinese characters are more suited for a handwritten input than a keyboard based one. Reason for that are the characters consisting mostly of pictograms or ideograms.
This thesis deals with an implementation of a prototypical recognition system on a mobile device. The recognition process should be online and therefore running while writing. It can save time for the user, because suggestions are made during runtime.
Basics and an overview over the current state of the art in online handwriting recognition will be given. An approach will be chosen and implemented, such that the recognition process is fast and needs little memory. The implementation will be tested and it will show, that a fast recognition can be possible on small devices. Suggestions for expansions and improvements will be given, including a future work part.
In den letzten Jahren ist eine steigende Verbreitung von Touchscreen-Geräten zu verzeichnen. Ihre Bedienung unterscheidet sich grundlegend von der mit Maus und Tastatur. Durch die Eingabe mit Gesten oder mehreren Fingern kann es schwierig sein den Aktionen eines Anderen zu folgen. Probleme entstehen durch die Verdeckung des Bildschirms mit der Eingabehand. Sieht man nur den Bildschirminhalt, zum Beispiel bei einer Videoübertragung, gehen Informationen über die Eingabe verloren.
In dieser Arbeit wird ein System entwickelt, das die kollaborative Arbeit an voneinander entfernten Touchscreen-Geräten verbessern soll. Dazu wird aus den Tiefendaten eines Kinect Sensors eine grafische Repräsentation der Eingabehand erstellt. Durch Einblendung dieser Visualisierung soll es einem Anwender erleichtert werden den Eingaben eines Anwenders zu folgen. Bedienkonzepte, wie zum Beispiel Gesten, sollen dadurch besser vermittelt werden. Außerdem soll so die Möglichkeit geschaffen werden, Informationen über eine gemeinsame Problematik effizienter auszutauschen. Deshalb wurde ein Testsystem mit zwei Arbeitsplätzen entwickelt. Darin übernimmt ein Anwender die Rolle des Erklärenden und führt einen zweiten Anwender, den Ausführenden, durch verschiedene Testszenarien. Hierbei stehen ihm bei einem Teil der Aufgaben die Visualisierung der Hand zur Verfügung, während er in anderen Aufgaben nur verbal mit seinem Gegenüber kommunizieren kann.
Im Rahmen einer Evaluation wird das System auf seine Effizienz zur Bedienung von Touchscreen-Systemen überprüft. Des Weiteren wird untersucht, inwieweit die grafische Qualität den gestellten Anforderungen genügt, um einen Mehrwert für die Anwendung zu bieten.
Tiny waves driven by wind, shallow, long waves, head overlapping sea, all of these waves occur in every ocean and even in small lakes. The surface of water is one of the most versatile phenomenas of nature. Not only the movement of waves, but also the reflection of sky, sun and coastline makes the surface of water unique. Exactly this complexity is what brings its own challenges to the simulation of water surfaces. That is why simulation of water occupies mathematicians with a challenge for nearly 400 years now.
In the last fifty years this challenge has more and more shifted to computer science. Computer graphic designers have tried to visualise water in a realistic manner for centuries. Science in this field expends from simple noise filters to mathematically complex solutions like Fourier Transformation.
In the following work historical background of todays wave theories, as well as mathematical fundamentals are given. The focus of this work is set on the implementation of these methods in OpenGL 3.3.
In this bachelor thesis, the question of whether or not a jump'n run game with sensor control for android devices is useful, is handled. To this end, a game was developed, which is once controlled with and without sensors at different levels. In a second version, the game is completely controlled by means of sensors, so that the controls can later be compared. It is explained how the game was planned, designed and investigated. In addition, it is checked whether games with sensor control already exist. The engine, which was used to developed the game, is also introduced. Finally, the evaluation is carried out for an elaborated user test on the playability of the game in terms of control.
The present thesis covers the implementation and optimization of global illumination in three-dimensional scenes. Global illumination does not only consider direct illumination dependent on one or more light sources, but also indirect illumination which is emitted by surrounding objects in the scene. The thesis focuses on the implementation of a global illumination method and its improvement using OpenGL 4.4. This is done by a voxelization of the scene. By traversing the resulting voxel structure, additional information is taken from the scene, which contributes to a plausible global illumination.
This thesis covers the mathematical background of ray-casting as well as an exemplary implementation on graphics processing units, using a modern programming interface. The implementation is embedded within an editor, which enables the user to activate optimizations of the algorithm. Techniques like transfer functions and local illumination are available for a more realistic visualization of materials. Moreover, the user interface gives access to features like importing volumes, let one define a custom transfer function, holds controls to adjust parameters of rendering and allows to activate further techniques, which are also subject of discussion in this thesis. Benefit of all shown techniques is measured, whether it is expected to be visual or on the part of performance.
Diese Arbeit beschreibt den Entwurf und die Umsetzung einer prototypischen Android-Applikation zum Dokumentieren von Reiserouten. Nach einer kurzen Einführung in die grundlegenden Aspekte des Reisens, der notwendigen Hilfsmittel sowie das Betriebssystem Android wird das Konzept und die Anforderungen an die Anwendung vorgestellt. Anschließend werden die Realisierung der Implementierung erläutert und die Ergebnisse der Evaluation offengelegt.
Ziel der Bachelorarbeit war es, eine moderne Art des Geländespiels "Schnitzeljagd" zu entwickeln. Dieses sollte möglichst auf jedem aktuellen Smartphone spielbar sein, welches mit dem Betriebssystem Android arbeitet. Das Gelände ist begrenzt auf den Universitätscampus Koblenz, somit dient das Spiel auch dazu, den Campus besser kennenzulernen.
Den Benutzern der Campusjagd wird eine mobile Applikation geboten, welche sie anhand von Hinweisen und Rätseln über den kompletten Campus führt, um letztendlich an einen Zielort zu gelangen, an welchem sich ein "Schatz" befindet. Anstatt wie üblich auf dem Gelände Schnitzel bzw. Schnipsel mit Hinweisen zu verstecken, werden bei der Campusjagd QR-Codes aufgehängt, um so den Weg zu markieren. Zur Täuschung sind auch irrelevante Codes zu finden. Die Codes müssen in der richtigen Reihenfolge abgelaufen werden, d. h. von einem Code erhält der Spieler den Hinweis zum jeweiligen Nächsten. Außerdem ist es möglich, dass aus einem QR-Code mehrere Hinweise für nächste Stationen angezeigt werden.
The goal of this work is evaluation and optimization of several eye-tracking algorithms for estimation of relevant features regarding accuracy. The extracted features are pupil- and glintcenters. The algorithms are applicable to off the shelf cameras. A synthetic model of the eye was modified and utilized. The model was used to supply ground truth for the evaluation of the methods.
Erweiterung der Konzeption und Implementierung einer Screening Applikation für mobile Endgeräte
(2014)
In this bachelor thesis an existing generic concept and an existing prototype for a smartphone application to record, monitor and document physical symptoms or observations of the human body are being extended. The existing funktionalities are being complemented by analysis of the previous Prototype. The concept and its Function modules, which are implemented in the existing prototype for the mobile platform Android, are being extended based on their analysed weaknesses. The resulting prototype and generic concept are evaluated and optimizations and extensions are being collected for further projects.
The animation of models has become an important part in different areas of everyday life. It is a demanding task for computer graphics to generate
a natural deformation of organic models. Skinning is a common method to animate models without animating each vertex individually. The skin of the model deforms automatically by manipulating individual bones of a skeleton.
This bachelor thesis deals with the most common algorithm, linear blend skinning and aims to find some optimizations regarding the visual effect
and performance. Additionally it presents certain instancing methods which are combined with the skinning methods in the application to show the advantages and disadvantages of the latter.
In the man-machine interaction tracking and identification of individuals plays an important role. In this work, a framework for the service-robot Lisa, of the Active Vision Group, has been created to combine different methods for the detection, tracking and identification of individuals. First leg detection is performed to establish hypotheses for people using a 2D-laserscan. This assumption needs to be confirmed by an analysis of the Kinect point cloud. After successful confirmation online-boosting on RGB-data is performed for identification. The leg data will also be used with a linear Kalman filter to estimate the movement of people. Through the combination of of Kalman filter with leg detection and online-boosting people tracking should be enabled. Further receiving an interchange of persons should - by brief occlusion or faulty associate of legs - can be prevented.
Abstract 3 This paper explains the convolution reverb, a method that enables users to add realistic sounding reverberation to audio material that was recorded in neutral sounding rooms. In particular, the possibility of computing the effect on the GPU using OpenCL is discussed, to make use of the high concurrency of the problem. This paper aims at the development of a VST plugin that utilizes the GPU accelerated convolution algorithm, so that it can be used for audio software solutions.
This work presents an application for simulation objects, which can change their aggregate states between solid and liquid using a temperature system. The focal points are the simulation of fluids with a particle system, the generation of a surface and the visualization of metal. The application should be interactive and match the real time conditions. Different types of Shader are used for the parallelized computations on the GPU. Also more options to use the application and possible improvements are presented.
This thesis shows an interaction of primitives in a three-dimensional space which is done by gestures. Functions which are difficult to do by gestures without any absolute feeling of the position are implemented with a touchscreen. Besides the touchscreen a second input device, a Leap-Motion, is used to obtain data of the motion of the hand. To get its data the Leap-Motion uses two CCD-cameras and three infrared LEDs. The interactions that can be done without any feedback of the absolute position are the translation, rotation and scale. These three and the movement through space are implemented as gestures in this thesis. This is done in Blender with the BlenderrnGame Engine and Python. The only function which has been implemented for the touchscreen is to select an object. Later on, a comparative control of the mouse was invented to contrast it with the control of the gestures. There are two big differences between these two controls. On the one hand, the gesture controls can be done in a three-dimensional space but most people aren't used to it yet. On the other hand, there is just a two-dimensional input possibility with the mouse control. Otherwise it is familiar to most persons. The evaluation should reveal if people prefer interaction by mouse control or by gestures. The result shows that the prefered control is done by the mouse. However in some categories of the tests the gestures are quite close to the result of the mouse.
Real-time graphics applications are tending to get more realistic and approximate real world illumination gets more reasonable due to improvement of graphics hardware. Using a wide variation of algorithms and ideas, graphics processing units (GPU) can simulate complex lighting situations rendering computer generated imagery with complicated effects such as shadows, refraction and reflection of light. Particularly, reflections are an improvement of realism, because they make shiny materials, e.g. brushed metals, wet surfaces like puddles or polished floors, appear more realistic and reveal information of their properties such as roughness and reflectance. Moreover, reflections can get more complex, depending on the view: a wet surface like a street during rain for example will reflect lights depending on the distance of the viewer, resulting in more streaky reflection, which will look more stretched, if the viewer is locatedrnfarther away from the light source. This bachelor thesis aims to give an overview of the state-of-the-art in terms of rendering reflections. Understanding light is a basic need to understand reflections and therefore a physical model of light and its reflection will be covered in section 2, followed by the motivational section 2.2, that will give visual appealing examples for reflections from the real world and the media. Coming to rendering techniques, first, the main principle will be explained in section 3 followed by a short general view of a wide variety of approaches that try to generate correct reflections in section 4. This thesis will describe the implementation of three major algorithms, that produce plausible local reflections. Therefore, the developed framework is described in section 5, then three major algorithms will be covered, that are common methods in most current game and graphics engines: Screen space reflections (SSR), parallax-corrected cube mapping (PCCM) and billboard reflections (BBR). After describing their functional principle, they will be analysed of their visual quality and the possibilities of their real-time application. Finally they will be compared to each other to investigate the advantages and disadvantages over each other. In conclusion, the gained experiences will be described by summarizing advantages and disadvantages of each technique and giving suggestions for improvements. A short perspective will be given, trying to create a view of upcoming real-time rendering techniques for the creation of reflections as specular effects.
In this thesis, an interactive application is developed for Android OS. The application is about a virtual-reality game. The game is settled in the genre of first-person shooters and takes place in a space scenario. By using a stereo renderer, it is possible to play the game combined with virtual-reality glasses.
Simulation von Schnee
(2015)
Physic simulations allow the creation of dynamic scenes on the computer. Computer generated images become lively and find use in movies, games and engineering applications. GPGPU techniques make use of the graphics card to simulate physics. The simulation of dynamic snow is still little researched. The Material Point Method is the first technique which is capable of showing the dynamics andrncharacteristics of snow.
The hybrid use of Lagrangian particles and a regular cartesian grid enables solving of partial differential equations. Therefore articles are transformed to the grid. The grid velocities can then be updated with the calculation of gradients in an FEM-manner (finite element method). Finally grid node velocities are weight back to the particles to move them across the scene. This method is coupled with a constitutive model to cover the dynamic nature of snow. This include collisions and breaking.
This bachelor thesis connects the recent developments in GPGPU techniques of OpenGL with the Material Point Method to efficiently simulate visually compelling, dynamic snow scenes.
Das Thema dieser Arbeit ist die Entwicklung einer hardwarebeschleunigten Einzelbildkompression zur Videoübertragung. Verfahren zur Einzelbildkompressionrn existieren bereits seit längerer Zeit. Jedoch genügen die gängigen Verfahren nicht den Anforderungen der Echtzeit und Performanz, um während einer Videoübertragung ohne spürbare Latenz zum Einsatz zu kommen. In dieser Arbeit soll einer der geläufigsten Algorithmen zur Bildkompression auf Parallelisierbarkeit, unter zu Hilfenahme der Grafikkarte, untersucht werden, um Echtzeitfähigkeit während der Kompression und Dekompression von computergenerierten Bildern zu erreichen. Die Ergebnisse werden evaluiert und in den Rahmen aktueller Verfahren parallelisierter Kompressionstechniken eingeordnet.
This thesis deals with the development of an interactive Android card game. As an example, the Hebrew game Yaniv was implemented. Focus is the elaboration of required background components and the corresponding implementation in that application. Required game processes will be screened and a possible solution will be identified.
For definite isolation and classification of important features in 3D multi-attribute volume data, multidimensional transfer functions are inalienable. Yet, when using multiple dimensions, the comprehension of the data and the interaction with it become a challenge. That- because neither the control of the versatile input parameters nor the visualization in a higher dimensional space are straightforward.
The goal of this thesis is the implementation of a transfer function editor which supports the creation of a multidimensional transfer function. Therefore different visualization and interaction techniques, like Parallel Coordinates, are used. Furthermore it will be possible to choose and combine the used dimensions interactively and the rendered volume will be adapted to the user interaction in real time.
Ray Tracing enables a close to reality rendering implementation of a modelled scene. Because of its functioning, it is able to display optical phenomena and complex lighting. Though, numerous computations per pixel have to be done. In practice implementations can not achieve computer graphics" aim of real-time rendering close to 60 frames per second. Current Graphics Processing Units (GPU) allows high execution parallelism of general-purpose computations. By using the graphics-API OpenGL this parallelism can be achieved and it is possible to design and realize a Ray-Tracer, which operates entirely on the GPU. The developed approach will be extended by an Uniform Grid - a Ray-Tracing acceleration structure. Hence, a speed-up is expected.
This thesis` purpose is the implementation of Ray-Tracer, which operates completely on the GPU, and its expansion by integrating an Uniform Grid. Afterwards, the evaluation of maximum achievable performance takes place. Possible problems regarding GPU-programming will be identified and analysed.
Thematik dieser Arbeit ist das dreidimensionale Image-Warping für diffuse und reflektierende Oberflächen. Das Warpingverfahren für den reflektierenden Fall gibt es erst seit 2014. Bei diesem neuen Algorithmus treten Artefakte auf, sobald ein Bild für einen alternativen Blickwinkel auf eine sehr unebene Fläche berechnet werden soll.
In dieser Arbeit wird der Weg von einem Raytracer, der die Eingabetexturen erzeugt, über das Warpingverfahren für beide Arten der Oberflächen, bis zur Optimierung des Reflective-Warping-Verfahrens erarbeitet. Schließlich werden die Ergebnisse der Optimierung bewertet und in den aktuellen sowie zukünftigen Stand der Technik eingeordnet.
The present work introduces a rigid-body physics engine, focusing on the collision detection by GPU. The increasing performance and accessibility of modern graphics cards ensures that they can be also used for algorithms that are meant not only for imaging. This advantage is used to implement an efficient collision detection based on particles. The performance differences between CPU and GPU are presented by using a test environment.
In this thesis we present an approach to track a RGB-D camera in 6DOF andconstruct 3D maps. We first acquire, register and synchronize RGB and depth images. After preprocessing we extract FAST features and match them between two consecutive frames. By depth projection we regain the z-value for the inlier correspondences. Afterwards we estimate the camera motion by 3D point set alignment between the correspondence set using least-squares. This local motion estimate is incrementally applied to a global transformation. Additionally wernpresent methods to build maps based on point cloud data acquired by a RGB-D camera. For map creation we use the OctoMap framework and optionally create a colored point cloud map. The system is evaluated with the widespread RGB-D benchmark.
This thesis presents an approach to optimizing the computation of soft shadows from area lights. The light source is sampled uniformly by traversing shadow rays as packets through an N-tree. This data structure stores an additional line space for every node. A line space stores precomputed information about geometry inside of shafts from one to another side of the node. This visibility information is used to terminate a ray. Additionally the graphics processing unit (short GPU) is used to speed up the computations through parallelism. The scene is rendered with OpenGL and the shadow value is computed on the GPU for each pixel. Evaluating the implementation shows a performance gain of 86% by comparison to the CPU, if using the GPU implementation. Using the line space instead of triangle intersections also increases the performance. The implementation provides good scaling with an increasing amount of triangles and has no visual disadvantages for many rays.
Augmented Reality has many areas of application. It can be used to simplify everyday life as well as working processes. However, since there are
many manufacturers that offer greatly varying systems, choosing the correct system according to application as well as cross-platform development are dfficult. This thesis attempts to develop an application which can be used to simulate Augmented Reality devices on Virtual Reality systems. This should simplify the processes of choosing a system as well as cross-platform
development.
Since the simulation will be designed to run on mobile devices, it should be possible to render high quality, realistic environments in advance, using a panoramic image. On a Virtual Reality device, they need to be displayed as a stereoscopic image. To achieve this, several methods are presented that can be used to perform this conversion. An editor will be created which will allow the creation of scenes, configuration of Augmented Reality devices and displaying them on a Virtual Reality system. For closing this thesis a test will be performed, to check the quality of the simulation as well as improvements that can be made.
A special challenge of the basic musical education of children is to give them an understanding of reading and writing musical scores. During the training of young choristers it is furthermore important to educate them in directly transforming the written scores into sounds.
Therefore it is an interesting idea to play the sounds to the children via piano or keyboard and simultaneously present them on a screen in musical notation.
The aim of this bachelor thesis is the implementation of such a system that allows to enter scores using a MIDI-compatible keyboard and then depicting these as musical notation. The prototype of the application operates in three steps. It receives the musical scores via keyboard in form of MIDIdatasets. These MIDI-information are converted to the MusicXML-format. Based on this MusicXML-notation the software finally generates and displays the visual output.
Der natürliche Prozess der Verwitterung ist ein komplexer Vorgang, der von unterschiedlichsten Parametern beeinflusstwird. Hauptbestandteil dieses Prozesses ist das Zusammenziehen des Fruchtvolumens infolge von Wasserverlust durch Transpiration sowie die Veränderung der Fruchtfarbe und Oberfläche. Es wurden bereits Verfahren entwickelt, die diese Eigenschaften mit Hilfe von Parametrisierung sowie physikalischer Ansätze simulieren. Die in dieser Arbeit erstellte Anwendung simuliert das Fruchtfleisch durch ein Tetraedernetz und die Veränderung der Haut mit Hilfe von dynamischer Texturanpassung. Der entwickelte Algorithmus arbeitet in linearer Laufzeit und seine Ergebnisse werden anhand selbst erstellter Fruchtmodelle präsentiert.
Die vorliegende Arbeit befasst sich mit der Entwicklung eines Pixel-Sticks, ein digitales Hilfsmittel für das Light Painting. Schwerpunkte sind die Entwicklung von Hard- und Software des Sticks, sowie einer Android Applikation für die Bedienung.
Der Einbezug des Smartphones unterscheidet diesen Pixel-Stick anderen Varianten. Neben der grundlegenden Steuerung ist es möglich, Bilder direkt von dem Smartphone auf den Stick zu übertragen.
This examination broaches the issue of an investigation that is supported by the eye-tracking technologie and uses the e-learn-system Compass. For this investigation, there was carried out an experiment to test participants if they recognize agreement errors in the subjekt-verb- congruence in german sentences. The idea for that came from an other survey, that was hold for the englisch language. The integration of the e-learn-system Compass intended the observation of the participants learning behaviour. The experiment was able to be evaluated and analysed by the software of the eye-tracker, that recorded all eye movements and created valuable data out of them. It was assumed to catch the participants eye movements such as fixations or regressions.
The result was very clear: errors in congruence betweet subject and verb were done often. Besides, the error-rate increases proportionally to the length of the sentence.
Virtueller Dirigent
(2016)
Ihm Rahmen einer Bachelorarbeit wurde eine "Virtual Reality"-Anwendung entwickelt. Ein Benuzter erhält die Möglichkeit die Rolle eines Dirigenten zu übernehmen und ein virtuelles Orchester zu leiten. Die Leitung erfolgt mit Hilfe von Handgesten. Mithilfe einer Evaluation der Anwendung wurden allgemeine Anforderungen an "Virtual Reality"-Spiele festgehalten.
Augmented reality is being present for many years. Through progress in technology smaller augmented reality glasses became possible. These new technologies allow many new ways of interaction and usage of augmented reality.
This thesis is about the Microsoft HoloLens and its possiblities for consumers and industry. In the context of this thesis a new interactive and augmented application to measure the possiblities and limitations of the Microsoft HoloLens has been developed. The scene is an assembly szenario with a step by step instruction of building with Lego bricks. The evaluation showed that the HoloLens can already be used to assist in assembling scenarios and offers some advantages over other methods, although the glasses still have some flaws.
This thesis presents two methods for the computation of global illumination. The first is an extension of Reflective Shadow Maps with an additional shadow test in order to handle occlusion. The second method is a novel, bidirectional Light-Injection approach. Rays originating from the light source are traced through the scene and stored inside the shafts of the Linespace datastructure. These shafts are a discretization of the possible spatial directions. The Linespaces are embedded in a Uniform Grid. When retrieving this pre-calculated lightning information no traversal of datastructures and no additional indirection is necessary in the best-case scenario. This reduces computation time and variance compared to Pathtracing. Areas that are mostly lit indirectly and glas profit the most from this. However, the result is only approximative in nature and produces visible artifacts.
This thesis tests several methods and measures in pathtracing for selecting either the Line Space or the Bounding Volume Hierarchy data structure to make use of the advantages of both. The structures are defined locally around each object and each Line Space shaft contains one candidate ID each. All implementation is done as a C++ and OpenGL framework with compute shaders handling the pathtracing and Line Space generation. The measures include the probability distribution, the effect dependency, as well as a distance threshold and are tested against several different scenes. In most situations, the results show a noticeable increase in performance, partly only with minor visual differences, with the probability measure producing the highest quality images for a given performance. The fundamental problems of the Line Space concering the high memory consumption and a long generation time compared to the BVH still persist, despite the object local structure, a minimal amount of data per shaft and the compute shader implementation.
This thesis presents a novel technique in computer graphics to simulate realtime
global illumination using path tracing. Path tracing is done with compute shaders on the graphics card (GPU) to perform rendering in a highly parallelized manner. To improve the overall performance of tracing rays, the Line Space is used as an acceleration data structure in different variations, resulting in better
empty space skipping. The Line Space saves scene information based on a previous voxelization in direction-dependent shafts and is generated and traversed on the GPU. With this procedure, indirect lighting and soft shadows can be computed in a physically correct way. Furthermore, using the Line Space, path tracing can be performed mostly independent of the complexity of the scene geometry with over 100 frames per second, which is truly real-time and much faster than using a comparable voxel grid. The image quality is not affected negatively by this technique and the shadow quality is in most cases much better compared to shadow-mapping.
Part-of-Speech tagging is the process of assigning words with similar grammatical properties to a part of speech (PoS). In the English language, PoS-tagging algorithms generally reach very high accuracy. This thesis undertakes the task to test against these accuracies in PoS-tagging as a qualitative measure in classification capabilities for a recently developed neural network model, called graph convolutional network (GCN). The novelty proposed in this thesis is to translate a corpus into a graph as a direct input for the GCN. The experiments in this thesis serve as a proof of concept with room for improvements.
This thesis deals with the exploration of different interaction possibilities
for three-dimensional, virtual objects in a real environment. The focus lies
especially on interaction possibilities from new AR-technologies.
A playful prototype of an application for Microsofts HoloLens will be
designed and implemented. The prototype consists of three parts. The first
part is the scan-process of the real environment of the user. In the second
part the user can augment the real environment with three-dimensional,
virtual objects. In the third part the user is supposed to navigate a virtual
avatar through the real environment.
The interaction possibilities of the HoloLens like Gaze, Gesture and VoiceInput
will be evaluated in the following categories menu navigation, positioning
of three-dimensional objects in a real environment and controlling an
avatar.
This thesis explores different approaches for the acceleration of raytracing calculations on the graphics processing unit (GPU). For that a voxel grid is used and extended by the linespace data structure. The linespace consists of direction based shafts and stores the objects located in those shafts in a candidate list. Different methods for the sorting and traversal of the linespace are presented and evaluated. The shown methods cannot provide a speed up of the frame rate without resulting in a loss of image quality.