@phdthesis{Wick2020,
  author    = {Wick, Christoph},
  title     = {Optical Medieval Music Recognition},
  doi       = {10.25972/OPUS-21434},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-214348},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2020},
  abstract  = {In recent years, great progress has been made in the area of Artificial Intelligence (AI) due to the possibilities of Deep Learning which steadily yielded new state-of-the-art results especially in many image recognition tasks. Currently, in some areas, human performance is achieved or already exceeded. This great development already had an impact on the area of Optical Music Recognition (OMR) as several novel methods relying on Deep Learning succeeded in specific tasks. Musicologists are interested in large-scale musical analysis and in publishing digital transcriptions in a collection enabling to develop tools for searching and data retrieving. The application of OMR promises to simplify and thus speed-up the transcription process by either providing fully-automatic or semi-automatic approaches. This thesis focuses on the automatic transcription of Medieval music with a focus on square notation which poses a challenging task due to complex layouts, highly varying handwritten notations, and degradation. However, since handwritten music notations are quite complex to read, even for an experienced musicologist, it is to be expected that even with new techniques of OMR manual corrections are required to obtain the transcriptions. This thesis presents several new approaches and open source software solutions for layout analysis and Automatic Text Recognition (ATR) for early documents and for OMR of Medieval manuscripts providing state-of-the-art technology. Fully Convolutional Networks (FCN) are applied for the segmentation of historical manuscripts and early printed books, to detect staff lines, and to recognize neume notations. The ATR engine Calamari is presented which allows for ATR of early prints and also the recognition of lyrics. Configurable CNN/LSTM-network architectures which are trained with the segmentation-free CTC-loss are applied to the sequential recognition of text but also monophonic music. Finally, a syllable-to-neume assignment algorithm is presented which represents the final step to obtain a complete transcription of the music. The evaluations show that the performances of any algorithm is highly depending on the material at hand and the number of training instances. The presented staff line detection correctly identifies staff lines and staves with an \$F_1\$-score of above \$99.5\\%\$. The symbol recognition yields a diplomatic Symbol Accuracy Rate (dSAR) of above \$90\\%\$ by counting the number of correct predictions in the symbols sequence normalized by its length. The ATR of lyrics achieved a Character Error Rate (CAR) (equivalently the number of correct predictions normalized by the sentence length) of above \$93\\%\$ trained on 771 lyric lines of Medieval manuscripts and of 99.89\\% when training on around 3.5 million lines of contemporary printed fonts. The assignment of syllables and their corresponding neumes reached \$F_1\$-scores of up to \$99.2\\%\$. A direct comparison to previously published performances is difficult due to different materials and metrics. However, estimations show that the reported values of this thesis exceed the state-of-the-art in the area of square notation. A further goal of this thesis is to enable musicologists without technical background to apply the developed algorithms in a complete workflow by providing a user-friendly and comfortable Graphical User Interface (GUI) encapsulating the technical details. For this purpose, this thesis presents the web-application OMMR4all. Its fully-functional workflow includes the proposed state-of-the-art machine-learning algorithms and optionally allows for a manual intervention at any stage to correct the output preventing error propagation. To simplify the manual (post-) correction, OMMR4all provides an overlay-editor that superimposes the annotations with a scan of the original manuscripts so that errors can easily be spotted. The workflow is designed to be iteratively improvable by training better models as soon as new Ground Truth (GT) is available.},
  subject      = {Neumenschrift},
  language  = {en}
}
@phdthesis{Borrmann2018,
  author    = {Borrmann, Dorit},
  title     = {Multi-modal 3D mapping - Combining 3D point clouds with thermal and color information},
  isbn      = {978-3-945459-20-1},
  issn      = {1868-7474},
  doi       = {10.25972/OPUS-15708},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-157085},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2018},
  abstract  = {Imagine a technology that automatically creates a full 3D thermal model of an environment and detects temperature peaks in it. For better orientation in the model it is enhanced with color information. The current state of the art for analyzing temperature related issues is thermal imaging. It is relevant for energy efficiency but also for securing important infrastructure such as power supplies and temperature regulation systems. Monitoring and analysis of the data for a large building is tedious as stable conditions need to be guaranteed for several hours and detailed notes about the pose and the environment conditions for each image must be taken. For some applications repeated measurements are necessary to monitor changes over time. The analysis of the scene is only possible through expertise and experience. This thesis proposes a robotic system that creates a full 3D model of the environment with color and thermal information by combining thermal imaging with the technology of terrestrial laser scanning. The addition of a color camera facilitates the interpretation of the data and allows for other application areas. The data from all sensors collected at different positions is joined in one common reference frame using calibration and scan matching. The first part of the thesis deals with 3D point cloud processing with the emphasis on accessing point cloud data efficiently, detecting planar structures in the data and registering multiple point clouds into one common coordinate system. The second part covers the autonomous exploration and data acquisition with a mobile robot with the objective to minimize the unseen area in 3D space. Furthermore, the combination of different modalities, color images, thermal images and point cloud data through calibration is elaborated. The last part presents applications for the the collected data. Among these are methods to detect the structure of building interiors for reconstruction purposes and subsequent detection and classification of windows. A system to project the gathered thermal information back into the scene is presented as well as methods to improve the color information and to join separately acquired point clouds and photo series. A full multi-modal 3D model contains all the relevant geometric information about the recorded scene and enables an expert to fully analyze it off-site. The technology clears the path for automatically detecting points of interest thereby helping the expert to analyze the heat flow as well as localize and identify heat leaks. The concept is modular and neither limited to achieving energy efficiency nor restricted to the use in combination with a mobile platform. It also finds its application in fields such as archaeology and geology and can be extended by further sensors.},
  subject      = {Punktwolke},
  language  = {en}
}
@phdthesis{Houshiar2017,
  author    = {Houshiar, Hamidreza},
  title     = {Documentation and mapping with 3D point cloud processing},
  isbn      = {978-3-945459-14-0},
  doi       = {10.25972/OPUS-14449},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-144493},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2017},
  abstract  = {3D point clouds are a de facto standard for 3D documentation and modelling. The advances in laser scanning technology broadens the usability and access to 3D measurement systems. 3D point clouds are used in many disciplines such as robotics, 3D modelling, archeology and surveying. Scanners are able to acquire up to a million of points per second to represent the environment with a dense point cloud. This represents the captured environment with a very high degree of detail. The combination of laser scanning technology with photography adds color information to the point clouds. Thus the environment is represented more realistically. Full 3D models of environments, without any occlusion, require multiple scans. Merging point clouds is a challenging process. This thesis presents methods for point cloud registration based on the panorama images generated from the scans. Image representation of point clouds introduces 2D image processing methods to 3D point clouds. Several projection methods for the generation of panorama maps of point clouds are presented in this thesis. Additionally, methods for point cloud reduction and compression based on the panorama maps are proposed. Due to the large amounts of data generated from the 3D measurement systems these methods are necessary to improve the point cloud processing, transmission and archiving. This thesis introduces point cloud processing methods as a novel framework for the digitisation of archeological excavations. The framework replaces the conventional documentation methods for excavation sites. It employs point clouds for the generation of the digital documentation of an excavation with the help of an archeologist on-site. The 3D point cloud is used not only for data representation but also for analysis and knowledge generation. Finally, this thesis presents an autonomous indoor mobile mapping system. The mapping system focuses on the sensor placement planning method. Capturing a complete environment requires several scans. The sensor placement planning method solves for the minimum required scans to digitise large environments. Combining this method with a navigation system on a mobile robot platform enables it to acquire data fully autonomously. This thesis introduces a novel hole detection method for point clouds to detect obscured parts of a captured environment. The sensor placement planning method selects the next scan position with the most coverage of the obscured environment. This reduces the required number of scans. The navigation system on the robot platform consist of path planning, path following and obstacle avoidance. This guarantees the safe navigation of the mobile robot platform between the scan positions. The sensor placement planning method is designed as a stand alone process that could be used with a mobile robot platform for autonomous mapping of an environment or as an assistant tool for the surveyor on scanning projects.},
  subject      = {3D Punktwolke},
  language  = {en}
}
@phdthesis{SchauerMarinRodrigues2020,
  author    = {Schauer Marin Rodrigues, Johannes},
  title     = {Detecting Changes and Finding Collisions in 3D Point Clouds : Data Structures and Algorithms for Post-Processing Large Datasets},
  isbn      = {978-3-945459-32-4},
  doi       = {10.25972/OPUS-21428},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-214285},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2020},
  abstract  = {Affordable prices for 3D laser range finders and mature software solutions for registering multiple point clouds in a common coordinate system paved the way for new areas of application for 3D point clouds. Nowadays we see 3D laser scanners being used not only by digital surveying experts but also by law enforcement officials, construction workers or archaeologists. Whether the purpose is digitizing factory production lines, preserving historic sites as digital heritage or recording environments for gaming or virtual reality applications -- it is hard to imagine a scenario in which the final point cloud must also contain the points of "moving" objects like factory workers, pedestrians, cars or flocks of birds. For most post-processing tasks, moving objects are undesirable not least because moving objects will appear in scans multiple times or are distorted due to their motion relative to the scanner rotation. The main contributions of this work are two postprocessing steps for already registered 3D point clouds. The first method is a new change detection approach based on a voxel grid which allows partitioning the input points into static and dynamic points using explicit change detection and subsequently remove the latter for a "cleaned" point cloud. The second method uses this cleaned point cloud as input for detecting collisions between points of the environment point cloud and a point cloud of a model that is moved through the scene. Our approach on explicit change detection is compared to the state of the art using multiple datasets including the popular KITTI dataset. We show how our solution achieves similar or better F1-scores than an existing solution while at the same time being faster. To detect collisions we do not produce a mesh but approximate the raw point cloud data by spheres or cylindrical volumes. We show how our data structures allow efficient nearest neighbor queries that make our CPU-only approach comparable to a massively-parallel algorithm running on a GPU. The utilized algorithms and data structures are discussed in detail. All our software is freely available for download under the terms of the GNU General Public license. Most of the datasets used in this thesis are freely available as well. We provide shell scripts that allow one to directly reproduce the quantitative results shown in this thesis for easy verification of our findings.},
  subject      = {Punktwolke},
  language  = {en}
}
@phdthesis{Leutert2021,
  author    = {Leutert, Florian},
  title     = {Flexible Augmented Reality Systeme f{\"u}r robotergest{\"u}tzte Produktionsumgebungen},
  isbn      = {978-3-945459-39-3},
  doi       = {10.25972/OPUS-24972},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-249728},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2021},
  abstract  = {Produktionssysteme mit Industrierobotern werden zunehmend komplex; waren deren Arbeitsbereiche fr{\"u}her noch statisch und abgeschirmt, und die programmierten Abl{\"a}ufe gleichbleibend, so sind die Anforderungen an moderne Robotik-Produktionsanlagen gestiegen: Diese sollen sich jetzt mithilfe von intelligenter Sensorik auch in unstrukturierten Umgebungen einsetzen lassen, sich bei sinkenden Losgr{\"o}ßen aufgrund individualisierter Produkte und h{\"a}ufig {\"a}ndernden Produktionsaufgaben leicht rekonfigurieren lassen, und sogar eine direkte Zusammenarbeit zwischen Mensch und Roboter erm{\"o}glichen. Gerade auch bei dieser Mensch-Roboter-Kollaboration wird es damit notwendig, dass der Mensch die Daten und Aktionen des Roboters leicht verstehen kann. Aufgrund der gestiegenen Anforderungen m{\"u}ssen somit auch die Bedienerschnittstellen dieser Systeme verbessert werden. Als Grundlage f{\"u}r diese neuen Benutzerschnittstellen bietet sich Augmented Reality (AR) als eine Technologie an, mit der sich komplexe r{\"a}umliche Daten f{\"u}r den Bediener leicht verst{\"a}ndlich darstellen lassen. Komplexe Informationen werden dabei in der Arbeitsumgebung der Nutzer visualisiert und als virtuelle Einblendungen sichtbar gemacht, und so auf einen Blick verst{\"a}ndlich. Die diversen existierenden AR-Anzeigetechniken sind f{\"u}r verschiedene Anwendungsfelder unterschiedlich gut geeignet, und sollten daher flexibel kombinier- und einsetzbar sein. Auch sollen diese AR-Systeme schnell und einfach auf verschiedenartiger Hardware in den unterschiedlichen Arbeitsumgebungen in Betrieb genommen werden k{\"o}nnen. In dieser Arbeit wird ein Framework f{\"u}r Augmented Reality Systeme vorgestellt, mit dem sich die genannten Anforderungen umsetzen lassen, ohne dass daf{\"u}r spezialisierte AR-Hardware notwendig wird. Das Flexible AR-Framework kombiniert und b{\"u}ndelt daf{\"u}r verschiedene Softwarefunktionen f{\"u}r die grundlegenden AR-Anzeigeberechnungen, f{\"u}r die Kalibrierung der notwendigen Hardware, Algorithmen zur Umgebungserfassung mittels Structured Light sowie generische ARVisualisierungen und erlaubt es dadurch, verschiedene AR-Anzeigesysteme schnell und flexibel in Betrieb zu nehmen und parallel zu betreiben. Im ersten Teil der Arbeit werden Standard-Hardware f{\"u}r verschiedene AR-Visualisierungsformen sowie die notwendigen Algorithmen vorgestellt, um diese flexibel zu einem AR-System zu kombinieren. Dabei m{\"u}ssen die einzelnen verwendeten Ger{\"a}te pr{\"a}zise kalibriert werden; hierf{\"u}r werden verschiedene M{\"o}glichkeiten vorgestellt, und die mit ihnen dann erreichbaren typischen Anzeige- Genauigkeiten in einer Evaluation charakterisiert. Nach der Vorstellung der grundlegenden ARSysteme des Flexiblen AR-Frameworks wird dann eine Reihe von Anwendungen vorgestellt, bei denen das entwickelte System in konkreten Praxis-Realisierungen als AR-Benutzerschnittstelle zum Einsatz kam, unter anderem zur {\"U}berwachung von, Zusammenarbeit mit und einfachen Programmierung von Industrierobotern, aber auch zur Visualisierung von komplexen Sensordaten oder zur Fernwartung. Im Verlauf der Arbeit werden dadurch die Vorteile, die sich durch Verwendung der AR-Technologie in komplexen Produktionssystemen ergeben, herausgearbeitet und in Nutzerstudien belegt.},
  subject      = {Erweiterte Realit{\"a}t <Informatik>},
  language  = {de}
}
@phdthesis{Strohmeier2021,
  author    = {Strohmeier, Michael},
  title     = {FARN - A Novel UAV Flight Controller for Highly Accurate and Reliable Navigation},
  doi       = {10.25972/OPUS-22313},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-223136},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2021},
  abstract  = {This thesis describes the functional principle of FARN, a novel flight controller for Unmanned Aerial Vehicles (UAVs) designed for mission scenarios that require highly accurate and reliable navigation. The required precision is achieved by combining low-cost inertial sensors and Ultra-Wide Band (UWB) radio ranging with raw and carrier phase observations from the Global Navigation Satellite System (GNSS). The flight controller is developed within the scope of this work regarding the mission requirements of two research projects, and successfully applied under real conditions. FARN includes a GNSS compass that allows a precise heading estimation even in environments where the conventional heading estimation based on a magnetic compass is not reliable. The GNSS compass combines the raw observations of two GNSS receivers with FARN's real-time capable attitude determination. Thus, especially the deployment of UAVs in Arctic environments within the project for ROBEX is possible despite the weak horizontal component of the Earth's magnetic field. Additionally, FARN allows centimeter-accurate relative positioning of multiple UAVs in real-time. This enables precise flight maneuvers within a swarm, but also the execution of cooperative tasks in which several UAVs have a common goal or are physically coupled. A drone defense system based on two cooperative drones that act in a coordinated manner and carry a commonly suspended net to capture a potentially dangerous drone in mid-air was developed in conjunction with the project MIDRAS. Within this thesis, both theoretical and practical aspects are covered regarding UAV development with an emphasis on the fields of signal processing, guidance and control, electrical engineering, robotics, computer science, and programming of embedded systems. Furthermore, this work aims to provide a condensed reference for further research in the field of UAVs. The work describes and models the utilized UAV platform, the propulsion system, the electronic design, and the utilized sensors. After establishing mathematical conventions for attitude representation, the actual core of the flight controller, namely the embedded ego-motion estimation and the principle control architecture are outlined. Subsequently, based on basic GNSS navigation algorithms, advanced carrier phase-based methods and their coupling to the ego-motion estimation framework are derived. Additionally, various implementation details and optimization steps of the system are described. The system is successfully deployed and tested within the two projects. After a critical examination and evaluation of the developed system, existing limitations and possible improvements are outlined.},
  subject      = {Drohne <Flugk{\"o}rper>},
  language  = {en}
}
@phdthesis{Pfitzner2019,
  author    = {Pfitzner, Christian},
  title     = {Visual Human Body Weight Estimation with Focus on Clinical Applications},
  isbn      = {978-3-945459-27-0 (online)},
  doi       = {10.25972/OPUS-17484},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-174842},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2019},
  abstract  = {It is the aim of this thesis to present a visual body weight estimation, which is suitable for medical applications. A typical scenario where the estimation of the body weight is essential, is the emergency treatment of stroke patients: In case of an ischemic stroke, the patient has to receive a body weight adapted drug, to solve a blood clot in a vessel. The accuracy of the estimated weight influences the outcome of the therapy directly. However, the treatment has to start as early as possible after the arrival at a trauma room, to provide sufficient treatment. Weighing a patient takes time, and the patient has to be moved. Furthermore, patients are often not able to communicate a value for their body weight due to their stroke symptoms. Therefore, it is state of the art that physicians guess the body weight. A patient receiving a too low dose has an increased risk that the blood clot does not dissolve and brain tissue is permanently damaged. Today, about one-third gets an insufficient dosage. In contrast to that, an overdose can cause bleedings and further complications. Physicians are aware of this issue, but a reliable alternative is missing. The thesis presents state-of-the-art principles and devices for the measurement and estimation of body weight in the context of medical applications. While scales are common and available at a hospital, the process of weighing takes too long and can hardly be integrated into the process of stroke treatment. Sensor systems and algorithms are presented in the section for related work and provide an overview of different approaches. The here presented system -- called Libra3D -- consists of a computer installed in a real trauma room, as well as visual sensors integrated into the ceiling. For the estimation of the body weight, the patient is on a stretcher which is placed in the field of view of the sensors. The three sensors -- two RGB-D and a thermal camera -- are calibrated intrinsically and extrinsically. Also, algorithms for sensor fusion are presented to align the data from all sensors which is the base for a reliable segmentation of the patient. A combination of state-of-the-art image and point cloud algorithms is used to localize the patient on the stretcher. The challenges in the scenario with the patient on the bed is the dynamic environment, including other people or medical devices in the field of view. After the successful segmentation, a set of hand-crafted features is extracted from the patient's point cloud. These features rely on geometric and statistical values and provide a robust input to a subsequent machine learning approach. The final estimation is done with a previously trained artificial neural network. The experiment section offers different configurations of the previously extracted feature vector. Additionally, the here presented approach is compared to state-of-the-art methods; the patient's own assessment, the physician's guess, and an anthropometric estimation. Besides the patient's own estimation, Libra3D outperforms all state-of-the-art estimation methods: 95 percent of all patients are estimated with a relative error of less than 10 percent to ground truth body weight. It takes only a minimal amount of time for the measurement, and the approach can easily be integrated into the treatment of stroke patients, while physicians are not hindered. Furthermore, the section for experiments demonstrates two additional applications: The extracted features can also be used to estimate the body weight of people standing, or even walking in front of a 3D camera. Also, it is possible to determine or classify the BMI of a subject on a stretcher. A potential application for this approach is the reduction of the radiation dose of patients being exposed to X-rays during a CT examination. During the time of this thesis, several data sets were recorded. These data sets contain the ground truth body weight, as well as the data from the sensors. They are available for the collaboration in the field of body weight estimation for medical applications.},
  subject      = {Punktwolke},
  language  = {en}
}
@phdthesis{Koch2018,
  author    = {Koch, Rainer},
  title     = {Sensor Fusion for Precise Mapping of Transparent and Specular Reflective Objects},
  isbn      = {978-3-945459-25-6},
  doi       = {10.25972/OPUS-16346},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-163462},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2018},
  abstract  = {Almost once a week broadcasts about earthquakes, hurricanes, tsunamis, or forest fires are filling the news. While oneself feels it is hard to watch such news, it is even harder for rescue troops to enter such areas. They need some skills to get a quick overview of the devastated area and find victims. Time is ticking, since the chance for survival shrinks the longer it takes till help is available. To coordinate the teams efficiently, all information needs to be collected at the command center. Therefore, teams investigate the destroyed houses and hollow spaces for victims. Doing so, they never can be sure that the building will not fully collapse while they are inside. Here, rescue robots are welcome helpers, as they are replaceable and make work more secure. Unfortunately, rescue robots are not usable off-the-shelf, yet. There is no doubt, that such a robot has to fulfil essential requirements to successfully accomplish a rescue mission. Apart from the mechanical requirements it has to be able to build a 3D map of the environment. This is essential to navigate through rough terrain and fulfil manipulation tasks (e.g. open doors). To build a map and gather environmental information, robots are equipped with multiple sensors. Since laser scanners produce precise measurements and support a wide scanning range, they are common visual sensors utilized for mapping. Unfortunately, they produce erroneous measurements when scanning transparent (e.g. glass, transparent plastic) or specular reflective objects (e.g. mirror, shiny metal). It is understood that such objects can be everywhere and a pre-manipulation to prevent their influences is impossible. Using additional sensors also bear risks. The problem is that these objects are occasionally visible, based on the incident angle of the laser beam, the surface, and the type of object. Hence, for transparent objects, measurements might result from the object surface or objects behind it. For specular reflective objects, measurements might result from the object surface or a mirrored object. These mirrored objects are illustrated behind the surface which is wrong. To obtain a precise map, the surfaces need to be recognised and mapped reliably. Otherwise, the robot navigates into it and crashes. Further, points behind the surface should be identified and treated based on the object type. Points behind a transparent surface should remain as they represent real objects. In contrast, Points behind a specular reflective surface should be erased. To do so, the object type needs to be classified. Unfortunately, none of the current approaches is capable to fulfil these requirements. Therefore, the following thesis addresses this problem to detect transparent and specular reflective objects and to identify their influences. To give the reader a start up, the first chapters describe: the theoretical background concerning propagation of light; sensor systems applied for range measurements; mapping approaches used in this work; and the state-of-the-art concerning detection and identification of transparent and specular reflective objects. Afterwards, the Reflection-Identification-Approach, which is the core of subject thesis is presented. It describes 2D and a 3D implementation to detect and classify such objects. Both are available as ROS-nodes. In the next chapter, various experiments demonstrate the applicability and reliability of these nodes. It proves that transparent and specular reflective objects can be detected and classified. Therefore, a Pre- and Post-Filter module is required in 2D. In 3D, classification is possible solely with the Pre-Filter. This is due to the higher amount of measurements. An example shows that an updatable mapping module allows the robot navigation to rely on refined maps. Otherwise, two individual maps are build which require a fusion afterwards. Finally, the last chapter summarizes the results and proposes suggestions for future work.},
  subject      = {laserscanner},
  language  = {en}
}
@phdthesis{Wagner2023,
  author    = {Wagner, Jan Cetric},
  title     = {Maximalnetzplan zur reaktiven Steuerung von Produktionsabl{\"a}ufen},
  isbn      = {978-3-945459-43-0},
  doi       = {10.25972/OPUS-30545},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-305452},
  school      = {Universit{\"a}t W{\"u}rzburg},
  pages     = {182},
  year      = {2023},
  abstract  = {In produzierenden Unternehmen werden verschiedene Vorgehensweisen zur Planung, {\"U}berwachung und Steuerung von Produktionsabl{\"a}ufen eingesetzt. Einer dieser Methoden wird als Vorgangsknotennetzplantechnik bezeichnet. Die einzelnen Produktionsschritte werden als Knoten definiert und durch Pfeile miteinander verbunden. Die Pfeile stellen die Beziehungen der jeweiligen Vorg{\"a}nge zueinander und damit den Produktionsablauf dar. Diese Technik erlaubt den Anwendern einen umfassenden {\"U}berblick {\"u}ber die einzelnen Prozessrelationen. Zus{\"a}tzlich k{\"o}nnen mit ihr Vorgangszeiten und Produktfertigstellungszeiten ermittelt werden, wodurch eine ausf{\"u}hrliche Planung der Produktion erm{\"o}glicht wird. Ein Nachteil dieser Technik begr{\"u}ndet sich in der alleinigen Darstellung einer ausf{\"u}hrbaren Prozessabfolge. Im Falle eines St{\"o}rungseintritts mit der Folge eines nicht durchf{\"u}hrbaren Vorgangs muss von dem origin{\"a}ren Prozess abgewichen werden. Aufgrund dessen wird eine Neuplanung erforderlich. Es werden Alternativen f{\"u}r den gest{\"o}rten Vorgang ben{\"o}tigt, um eine Fortf{\"u}hrung des Prozesses ungeachtet der St{\"o}rung zu erreichen. Innerhalb dieser Arbeit wird daher eine Erweiterung der Vorgangsknotennetzplantechnik beschrieben, die es erlaubt, erg{\"a}nzend zu dem geplanten Soll-Prozess Alternativvorg{\"a}nge f{\"u}r einzelne Vorg{\"a}nge darzulegen. Diese Methode wird als Maximalnetzplan bezeichnet. Die Alternativen werden im Falle eines St{\"o}rungseintritts automatisch evaluiert und dem Anwender in priorisierter Reihenfolge pr{\"a}sentiert. Durch die Verwendung des Maximalnetzplans kann eine aufwendige Neuplanung vermieden werden. Als Anwendungsbeispiel dient ein Montageprozess, mithilfe dessen die Verwendbarkeit der Methode dargelegt wird. Weiterf{\"u}hrend zeigt eine zeitliche Analyse zufallsbedingter Maximalnetzpl{\"a}ne eine Begr{\"u}ndung zur Durchf{\"u}hrung von Alternativen und damit den Nutzen des Maximalnetzplans auf. Zus{\"a}tzlich sei angemerkt, dass innerhalb dieser Arbeit verwendete Begrifflichkeiten wie Anwender, Werker oder Mitarbeiter in maskuliner Schreibweise niedergeschrieben werden. Dieses ist ausschließlich der Einfachheit geschuldet und nicht dem Zweck der Diskriminierung anderer Geschlechter dienlich. Die verwendete Schreibweise soll alle Geschlechter ansprechen, ob m{\"a}nnlich, weiblich oder divers.},
  subject      = {Produktionsplanung},
  language  = {de}
}
@phdthesis{Sauer2023,
  author    = {Sauer, Christian},
  title     = {Development, Simulation and Evaluation of Mobile Wireless Networks in Industrial Applications},
  doi       = {10.25972/OPUS-29923},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-299238},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Manyindustrialautomationsolutionsusewirelesscommunicationandrelyontheavail- ability and quality of the wireless channel. At the same time the wireless medium is highly congested and guaranteeing the availability of wireless channels is becoming increasingly difficult. In this work we show, that ad-hoc networking solutions can be used to provide new communication channels and improve the performance of mobile automation systems. These ad-hoc networking solutions describe different communi- cation strategies, but avoid relying on network infrastructure by utilizing the Peer-to- Peer (P2P) channel between communicating entities. This work is a step towards the effective implementation of low-range communication technologies(e.g. VisibleLightCommunication(VLC), radarcommunication, mmWave communication) to the industrial application. Implementing infrastructure networks with these technologies is unrealistic, since the low communication range would neces- sitate a high number of Access Points (APs) to yield full coverage. However, ad-hoc networks do not require any network infrastructure. In this work different ad-hoc net- working solutions for the industrial use case are presented and tools and models for their examination are proposed. The main use case investigated in this work are Automated Guided Vehicles (AGVs) for industrial applications. These mobile devices drive throughout the factory trans- porting crates, goods or tools or assisting workers. In most implementations they must exchange data with a Central Control Unit (CCU) and between one another. Predicting if a certain communication technology is suitable for an application is very challenging since the applications and the resulting requirements are very heterogeneous. The proposed models and simulation tools enable the simulation of the complex inter- action of mobile robotic clients and a wireless communication network. The goal is to predict the characteristics of a networked AGV fleet. Theproposedtoolswereusedtoimplement, testandexaminedifferentad-hocnetwork- ing solutions for industrial applications using AGVs. These communication solutions handle time-critical and delay-tolerant communication. Additionally a control method for the AGVs is proposed, which optimizes the communication and in turn increases the transport performance of the AGV fleet. Therefore, this work provides not only tools for the further research of industrial ad-hoc system, but also first implementations of ad-hoc systems which address many of the most pressing issues in industrial applica- tions.},
  subject      = {Industrie},
  language  = {en}
}
@phdthesis{Bleier2023,
  author    = {Bleier, Michael},
  title     = {Underwater Laser Scanning - Refractive Calibration, Self-calibration and Mapping for 3D Reconstruction},
  isbn      = {978-3-945459-45-4},
  doi       = {10.25972/OPUS-32269},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-322693},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {There is great interest in affordable, precise and reliable metrology underwater: Archaeologists want to document artifacts in situ with high detail. In marine research, biologists require the tools to monitor coral growth and geologists need recordings to model sediment transport. Furthermore, for offshore construction projects, maintenance and inspection millimeter-accurate measurements of defects and offshore structures are essential. While the process of digitizing individual objects and complete sites on land is well understood and standard methods, such as Structure from Motion or terrestrial laser scanning, are regularly applied, precise underwater surveying with high resolution is still a complex and difficult task. Applying optical scanning techniques in water is challenging due to reduced visibility caused by turbidity and light absorption. However, optical underwater scanners provide significant advantages in terms of achievable resolution and accuracy compared to acoustic systems. This thesis proposes an underwater laser scanning system and the algorithms for creating dense and accurate 3D scans in water. It is based on laser triangulation and the main optical components are an underwater camera and a cross-line laser projector. The prototype is configured with a motorized yaw axis for capturing scans from a tripod. Alternatively, it is mounted to a moving platform for mobile mapping. The main focus lies on the refractive calibration of the underwater camera and laser projector, the image processing and 3D reconstruction. For highest accuracy, the refraction at the individual media interfaces must be taken into account. This is addressed by an optimization-based calibration framework using a physical-geometric camera model derived from an analytical formulation of a ray-tracing projection model. In addition to scanning underwater structures, this work presents the 3D acquisition of semi-submerged structures and the correction of refraction effects. As in-situ calibration in water is complex and time-consuming, the challenge of transferring an in-air scanner calibration to water without re-calibration is investigated, as well as self-calibration techniques for structured light. The system was successfully deployed in various configurations for both static scanning and mobile mapping. An evaluation of the calibration and 3D reconstruction using reference objects and a comparison of free-form surfaces in clear water demonstrate the high accuracy potential in the range of one millimeter to less than one centimeter, depending on the measurement distance. Mobile underwater mapping and motion compensation based on visual-inertial odometry is demonstrated using a new optical underwater scanner based on fringe projection. Continuous registration of individual scans allows the acquisition of 3D models from an underwater vehicle. RGB images captured in parallel are used to create 3D point clouds of underwater scenes in full color. 3D maps are useful to the operator during the remote control of underwater vehicles and provide the building blocks to enable offshore inspection and surveying tasks. The advancing automation of the measurement technology will allow non-experts to use it, significantly reduce acquisition time and increase accuracy, making underwater metrology more cost-effective.},
  subject      = {Selbstkalibrierung},
  language  = {en}
}
@phdthesis{Dhillon2023,
  author    = {Dhillon, Maninder Singh},
  title     = {Potential of Remote Sensing in Modeling Long-Term Crop Yields},
  doi       = {10.25972/OPUS-32258},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-322581},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Accurate crop monitoring in response to climate change at a regional or field scale plays a significant role in developing agricultural policies, improving food security, forecasting, and analysing global trade trends. Climate change is expected to significantly impact agriculture, with shifts in temperature, precipitation patterns, and extreme weather events negatively affecting crop yields, soil fertility, water availability, biodiversity, and crop growing conditions. Remote sensing (RS) can provide valuable information combined with crop growth models (CGMs) for yield assessment by monitoring crop development, detecting crop changes, and assessing the impact of climate change on crop yields. This dissertation aims to investigate the potential of RS data on modelling long-term crop yields of winter wheat (WW) and oil seed rape (OSR) for the Free State of Bavaria (70,550 km2 ), Germany. The first chapter of the dissertation describes the reasons favouring the importance of accurate crop yield predictions for achieving sustainability in agriculture. Chapter second explores the accuracy assessment of the synthetic RS data by fusing NDVIs of two high spatial resolution data (high pair) (Landsat (30 m, 16-days; L) and Sentinel-2 (10 m, 5-6 days; S), with four low spatial resolution data (low pair) (MOD13Q1 (250 m, 16-days), MCD43A4 (500 m, one day), MOD09GQ (250 m, one-day), and MOD09Q1 (250 m, 8-days)) using the spatial and temporal adaptive reflectance fusion model (STARFM), which fills regions' cloud or shadow gaps without losing spatial information. The chapter finds that both L-MOD13Q1 (R2 = 0.62, RMSE = 0.11) and S-MOD13Q1 (R2 = 0.68, RMSE = 0.13) are more suitable for agricultural monitoring than the other synthetic products fused. Chapter third explores the ability of the synthetic spatiotemporal datasets (obtained in chapter 2) to accurately map and monitor crop yields of WW and OSR at a regional scale. The chapter investigates and discusses the optimal spatial (10 m, 30 m, or 250 m), temporal (8 or 16-day) and CGMs (World Food Studies (WOFOST), and the semi-empiric light use efficiency approach (LUE)) for accurate crop yield estimations of both crop types. Chapter third observes that the observations of high temporal resolution (8-day) products of both S-MOD13Q1 and L-MOD13Q1 play a significant role in accurately measuring the yield of WW and OSR. The chapter investigates that the simple light use efficiency (LUE) model (R2 = 0.77 and relative RMSE (RRMSE) = 8.17\%) that required fewer input parameters to simulate crop yield is highly accurate, reliable, and more precise than the complex WOFOST model (R2 = 0.66 and RRMSE = 11.35\%) with higher input parameters. Chapter four researches the relationship of spatiotemporal fusion modelling using STRAFM on crop yield prediction for WW and OSR using the LUE model for Bavaria from 2001 to 2019. The chapter states the high positive correlation coefficient (R) = 0.81 and R = 0.77 between the yearly R2 of synthetic accuracy and modelled yield accuracy for WW and OSR from 2001 to 2019, respectively. The chapter analyses the impact of climate variables on crop yield predictions by observing an increase in R2 (0.79 (WW)/0.86 (OSR)) and a decrease in RMSE (4.51/2.57 dt/ha) when the climate effect is included in the model. The fifth chapter suggests that the coupling of the LUE model to the random forest (RF) model can further reduce the relative root mean square error (RRMSE) from -8\% (WW) and -1.6\% (OSR) and increase the R2 by 14.3\% (for both WW and OSR), compared to results just relying on LUE. The same chapter concludes that satellite-based crop biomass, solar radiation, and temperature are the most influential variables in the yield prediction of both crop types. Chapter six attempts to discuss both pros and cons of RS technology while analysing the impact of land use diversity on crop-modelled biomass of WW and OSR. The chapter finds that the modelled biomass of both crops is positively impacted by land use diversity to the radius of 450 (Shannon Diversity Index ~0.75) and 1050 m (~0.75), respectively. The chapter also discusses the future implications by stating that including some dependent factors (such as the management practices used, soil health, pest management, and pollinators) could improve the relationship of RS-modelled crop yields with biodiversity. Lastly, chapter seven discusses testing the scope of new sensors such as unmanned aerial vehicles, hyperspectral sensors, or Sentinel-1 SAR in RS for achieving accurate crop yield predictions for precision farming. In addition, the chapter highlights the significance of artificial intelligence (AI) or deep learning (DL) in obtaining higher crop yield accuracies.},
  subject      = {Ernteertrag},
  language  = {en}
}
@phdthesis{Dhillon2023,
  author    = {Dhillon, Maninder Singh},
  title     = {Potential of Remote Sensing in Modeling Long-Term Crop Yields},
  doi       = {10.25972/OPUS-33052},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-330529},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Accurate crop monitoring in response to climate change at a regional or field scale plays a significant role in developing agricultural policies, improving food security, forecasting, and analysing global trade trends. Climate change is expected to significantly impact agriculture, with shifts in temperature, precipitation patterns, and extreme weather events negatively affecting crop yields, soil fertility, water availability, biodiversity, and crop growing conditions. Remote sensing (RS) can provide valuable information combined with crop growth models (CGMs) for yield assessment by monitoring crop development, detecting crop changes, and assessing the impact of climate change on crop yields. This dissertation aims to investigate the potential of RS data on modelling long-term crop yields of winter wheat (WW) and oil seed rape (OSR) for the Free State of Bavaria (70,550 km2), Germany. The first chapter of the dissertation describes the reasons favouring the importance of accurate crop yield predictions for achieving sustainability in agriculture. Chapter second explores the accuracy assessment of the synthetic RS data by fusing NDVIs of two high spatial resolution data (high pair) (Landsat (30 m, 16-days; L) and Sentinel-2 (10 m, 5-6 days; S), with four low spatial resolution data (low pair) (MOD13Q1 (250 m, 16-days), MCD43A4 (500 m, one day), MOD09GQ (250 m, one-day), and MOD09Q1 (250 m, 8-days)) using the spatial and temporal adaptive reflectance fusion model (STARFM), which fills regions' cloud or shadow gaps without losing spatial information. The chapter finds that both L-MOD13Q1 (R2 = 0.62, RMSE = 0.11) and S-MOD13Q1 (R2 = 0.68, RMSE = 0.13) are more suitable for agricultural monitoring than the other synthetic products fused. Chapter third explores the ability of the synthetic spatiotemporal datasets (obtained in chapter 2) to accurately map and monitor crop yields of WW and OSR at a regional scale. The chapter investigates and discusses the optimal spatial (10 m, 30 m, or 250 m), temporal (8 or 16-day) and CGMs (World Food Studies (WOFOST), and the semi-empiric light use efficiency approach (LUE)) for accurate crop yield estimations of both crop types. Chapter third observes that the observations of high temporal resolution (8-day) products of both S-MOD13Q1 and L-MOD13Q1 play a significant role in accurately measuring the yield of WW and OSR. The chapter investigates that the simple light use efficiency (LUE) model (R2 = 0.77 and relative RMSE (RRMSE) = 8.17\%) that required fewer input parameters to simulate crop yield is highly accurate, reliable, and more precise than the complex WOFOST model (R2 = 0.66 and RRMSE = 11.35\%) with higher input parameters. Chapter four researches the relationship of spatiotemporal fusion modelling using STRAFM on crop yield prediction for WW and OSR using the LUE model for Bavaria from 2001 to 2019. The chapter states the high positive correlation coefficient (R) = 0.81 and R = 0.77 between the yearly R2 of synthetic accuracy and modelled yield accuracy for WW and OSR from 2001 to 2019, respectively. The chapter analyses the impact of climate variables on crop yield predictions by observing an increase in R2 (0.79 (WW)/0.86 (OSR)) and a decrease in RMSE (4.51/2.57 dt/ha) when the climate effect is included in the model. The fifth chapter suggests that the coupling of the LUE model to the random forest (RF) model can further reduce the relative root mean square error (RRMSE) from -8\% (WW) and -1.6\% (OSR) and increase the R2 by 14.3\% (for both WW and OSR), compared to results just relying on LUE. The same chapter concludes that satellite-based crop biomass, solar radiation, and temperature are the most influential variables in the yield prediction of both crop types. Chapter six attempts to discuss both pros and cons of RS technology while analysing the impact of land use diversity on crop-modelled biomass of WW and OSR. The chapter finds that the modelled biomass of both crops is positively impacted by land use diversity to the radius of 450 (Shannon Diversity Index ~0.75) and 1050 m (~0.75), respectively. The chapter also discusses the future implications by stating that including some dependent factors (such as the management practices used, soil health, pest management, and pollinators) could improve the relationship of RS-modelled crop yields with biodiversity. Lastly, chapter seven discusses testing the scope of new sensors such as unmanned aerial vehicles, hyperspectral sensors, or Sentinel-1 SAR in RS for achieving accurate crop yield predictions for precision farming. In addition, the chapter highlights the significance of artificial intelligence (AI) or deep learning (DL) in obtaining higher crop yield accuracies.},
  subject      = {Ernteertrag},
  language  = {en}
}
@phdthesis{Krenzer2023,
  author    = {Krenzer, Adrian},
  title     = {Machine learning to support physicians in endoscopic examinations with a focus on automatic polyp detection in images and videos},
  doi       = {10.25972/OPUS-31911},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-319119},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Deep learning enables enormous progress in many computer vision-related tasks. Artificial Intel- ligence (AI) steadily yields new state-of-the-art results in the field of detection and classification. Thereby AI performance equals or exceeds human performance. Those achievements impacted many domains, including medical applications. One particular field of medical applications is gastroenterology. In gastroenterology, machine learning algorithms are used to assist examiners during interventions. One of the most critical concerns for gastroenterologists is the development of Colorectal Cancer (CRC), which is one of the leading causes of cancer-related deaths worldwide. Detecting polyps in screening colonoscopies is the essential procedure to prevent CRC. Thereby, the gastroenterologist uses an endoscope to screen the whole colon to find polyps during a colonoscopy. Polyps are mucosal growths that can vary in severity. This thesis supports gastroenterologists in their examinations with automated detection and clas- sification systems for polyps. The main contribution is a real-time polyp detection system. This system is ready to be installed in any gastroenterology practice worldwide using open-source soft- ware. The system achieves state-of-the-art detection results and is currently evaluated in a clinical trial in four different centers in Germany. The thesis presents two additional key contributions: One is a polyp detection system with ex- tended vision tested in an animal trial. Polyps often hide behind folds or in uninvestigated areas. Therefore, the polyp detection system with extended vision uses an endoscope assisted by two additional cameras to see behind those folds. If a polyp is detected, the endoscopist receives a vi- sual signal. While the detection system handles the additional two camera inputs, the endoscopist focuses on the main camera as usual. The second one are two polyp classification models, one for the classification based on shape (Paris) and the other on surface and texture (NBI International Colorectal Endoscopic (NICE) classification). Both classifications help the endoscopist with the treatment of and the decisions about the detected polyp. The key algorithms of the thesis achieve state-of-the-art performance. Outstandingly, the polyp detection system tested on a highly demanding video data set shows an F1 score of 90.25 \% while working in real-time. The results exceed all real-time systems in the literature. Furthermore, the first preliminary results of the clinical trial of the polyp detection system suggest a high Adenoma Detection Rate (ADR). In the preliminary study, all polyps were detected by the polyp detection system, and the system achieved a high usability score of 96.3 (max 100). The Paris classification model achieved an F1 score of 89.35 \% which is state-of-the-art. The NICE classification model achieved an F1 score of 81.13 \%. Furthermore, a large data set for polyp detection and classification was created during this thesis. Therefore a fast and robust annotation system called Fast Colonoscopy Annotation Tool (FastCAT) was developed. The system simplifies the annotation process for gastroenterologists. Thereby the i gastroenterologists only annotate key parts of the endoscopic video. Afterward, those video parts are pre-labeled by a polyp detection AI to speed up the process. After the AI has pre-labeled the frames, non-experts correct and finish the annotation. This annotation process is fast and ensures high quality. FastCAT reduces the overall workload of the gastroenterologist on average by a factor of 20 compared to an open-source state-of-art annotation tool.},
  subject      = {Deep Learning},
  language  = {en}
}