Der Wettbewerb um die besten Technologien zur Realisierung des autonomen Fahrens ist weltweit in vollem Gange. Trotz großer Anstrengungen ist jedoch die autonome Navigation in strukturierter und vor allem unstrukturierter Umgebung bisher nicht gelöst. Ein entscheidender Baustein in diesem Themenkomplex ist die Umgebungswahrnehmung und Analyse durch passende Sensorik und entsprechende Sensordatenauswertung. Insbesondere bildgebende Verfahren im Bereich des für den Menschen sichtbaren Spektrums finden sowohl in der Praxis als auch in der Forschung breite Anwendung. Dadurch wird jedoch nur ein Bruchteil des elektromagnetischen Spektrums genutzt und folglich ein großer Teil der verfügbaren Informationen zur Umgebungswahrnehmung ignoriert. Um das vorhandene Spektrum besser zu nutzen, werden in anderen Forschungsbereichen schon seit Jahrzehnten \sog spektrale Sensoren eingesetzt, welche das elektromagnetische Spektrum wesentlich feiner und in einem größeren Bereich im Vergleich zu klassischen Farbkameras analysieren. Jedoch können diese Systeme aufgrund technischer Limitationen nur statische Szenen aufnehmen. Neueste Entwicklungen der Sensortechnik ermöglichen nun dank der \sog Snapshot-Mosaik-Filter-Technik die spektrale Abtastung dynamischer Szenen. In dieser Dissertation wird der Einsatz und die Eignung der Snapshot-Mosaik-Technik zur Umgebungswahrnehmung und Szenenanalyse im Bereich der autonomen Navigation in strukturierten und unstrukturierten Umgebungen untersucht. Dazu wird erforscht, ob die aufgenommen spektralen Daten einen Vorteil gegenüber klassischen RGB- \bzw Grauwertdaten hinsichtlich der semantischen Szenenanalyse und Klassifikation bieten. Zunächst wird eine geeignete Vorverarbeitung entwickelt, welche aus den Rohdaten der Sensorik spektrale Werte berechnet. Anschließend wird der Aufbau von neuartigen Datensätzen mit spektralen Daten erläutert. Diese Datensätze dienen als Basis zur Evaluation von verschiedenen Klassifikatoren aus dem Bereich des klassischen maschinellen Lernens. Darauf aufbauend werden Methoden und Architekturen aus dem Bereich des Deep-Learnings vorgestellt. Anhand ausgewählter Architekturen wird untersucht, ob diese auch mit spektralen Daten trainiert werden können. Weiterhin wird die Verwendung von Deep-Learning-Methoden zur Datenkompression thematisiert. In einem nächsten Schritt werden die komprimierten Daten genutzt, um damit Netzarchitekturen zu trainieren, welche bisher nur mit RGB-Daten kompatibel sind. Abschließend wird analysiert, ob die hochdimensionalen spektralen Daten bei der Szenenanalyse Vorteile gegenüber RGB-Daten bieten

Habitat loss and fragmentation due to climate and land-use change are among the biggest threats to biodiversity, as the survival of species relies on suitable habitat area and the possibility to disperse between different patches of habitat. To predict and mitigate the effects of habitat loss, a better understanding of species dispersal is needed. Graph theory provides powerful tools to model metapopulations in changing landscapes with the help of habitat networks, where nodes represent habitat patches and links indicate the possible dispersal pathways between patches. This thesis adapts tools from graph theory and optimisation to study species dispersal on habitat networks as well as the structure of habitat networks and the effects of habitat loss. In chapter 1, I will give an introduction to the thesis and the different topics presented in this thesis. Chapter 2 will then give a brief summary of tools used in the thesis. In chapter 3, I present our model on possible range shifts for a generic species. Based on a graph-based dispersal model for a generic aquatic invertebrate with a terrestrial life stage, we developed an optimisation model that models dispersal directed to predefined habitat patches and yields a minimum time until these patches are colonised with respect to the given landscape structure and species dispersal capabilities. We created a time-expanded network based on the original habitat network and solved a mixed integer program to obtain the minimum colonisation time. The results provide maximum possible range shifts, and can be used to estimate how fast newly formed habitat patches can be colonised. Although being specific for this simulation model, the general idea of deriving a surrogate can in principle be adapted to other simulation models. Next, in chapter 4, I present our model to evaluate the robustness of metapopulations. Based on a variety of habitat networks and different generic species characterised by their dispersal traits and habitat demands, we modeled the permanent loss of habitat patches and subsequent metapopulation dynamics. The results show that species with short dispersal ranges and high local-extinction risks are particularly vulnerable to the loss of habitat across all types of networks. On this basis, we then investigated how well different graph-theoretic metrics of habitat networks can serve as indicators of metapopulation robustness against habitat loss. We identified the clustering coefficient of a network as the only good proxy for metapopulation robustness across all types of species, networks, and habitat loss scenarios. Finally, in chapter 5, I utilise the results obtained in chapter 4 to identify the areas in a network that should be improved in terms of restoration to maximise the metapopulation robustness under limited resources. More specifically, we exploit our findings that a network’s clustering coefficient is a good indicator for metapopulation robustness and develop two heuristics, a Greedy algorithm and a deducted Lazy Greedy algorithm, that aim at maximising the clustering coefficient of a network. Both algorithms can be applied to any network and are not specific to habitat networks only. In chapter 6, I will summarize the main findings of this thesis, discuss their limitations and give an outlook of future research topics. Overall this thesis develops frameworks to study the behaviour of habitat networks and introduces mathematical tools to ecology and thus narrows the gap between mathematics and ecology. While all models in this thesis were developed with a focus on aquatic invertebrates, they can easily be adapted to other metapopulations.

Ray tracing acceleration through dedicated data structures has long been an important topic in computer graphics. In general, two different approaches are proposed: spatial and directional acceleration structures. The thesis at hand presents an innovative combined approach of these two areas, which enables a further acceleration of the tracing process of rays. State-of-the-art spatial data structures are used as base structures and enhanced by precomputed directional visibility information based on a sophisticated abstraction concept of shafts within an original structure, the Line Space. In the course of the work, novel approaches for the precomputed visibility information are proposed: a binary value that indicates whether a shaft is empty or non-empty as well as a single candidate approximating the actual surface as a representative candidate. It is shown how the binary value is used in a simple but effective empty space skipping technique, which allows a performance gain in ray tracing of up to 40% compared to the pure base data structure, regardless of the spatial structure that is actually used. In addition, it is shown that this binary visibility information provides a fast technique for calculating soft shadows and ambient occlusion based on blocker approximations. Although the results contain a certain inaccuracy error, which is also presented and discussed, it is shown that a further tracing acceleration of up to 300% compared to the base structure is achieved. As an extension of this approach, the representative candidate precomputation is demonstrated, which is used to accelerate the indirect lighting computation, resulting in a significant performance gain at the expense of image errors. Finally, techniques based on two-stage structures and a usage heuristic are proposed and evaluated. These reduce memory consumption and approximation errors while maintaining the performance gain and also enabling further possibilities with object instancing and rigid transformations. All performance and memory values as well as the approximation errors are measured, presented and discussed. Overall, the Line Space is shown to result in a considerate improvement in ray tracing performance at the cost of higher memory consumption and possible approximation errors. The presented findings thus demonstrate the capability of the combined approach and enable further possibilities for future work.

The sediments of surface waters are temporary or final depository of many chemical compounds, including trace metals and metalloids (metal(loid)s) from natural and anthropogenic sources. Whether they act as a source or sink of metal(loid)s depends strongly on the dynamics of the biogeochemical processes that take place at the sediment-water interface (SWI). Important information on biogeochemical processes as well as on the exposure, the fate and the transport of pollutants at the SWI can be obtained by determining chemical concentration profiles in the sediment pore water. A major challenge is to conduct experiments with a spatial resolution, which allows to adequately record existing gradients and to log all the parameters needed, to describe and better understand the complex processes at the SWIs. At the same time, it is from major importance to prevent the formation of any artifacts during sampling, which may occur due to the labile nature of the SWIs and the very steep biogeochemical gradients. In this context, in the first part of this work, a system was developed and tested that enables the automated, minimal invasive sampling of sediment pore water of undisturbed or manipulated sediments while simultaneously recording parameters such as redox potential, oxygen content and pH value. In an incubation experiment the impact of acidification and mechanical disturbance (re-suspension) on the mobility of 13 metal(loid)s was investigated using a triple quadrupole inductively coupled plasma-mass spectrometry (ICP-QQQ-MS) multi-element approach. Most metals were released as consequence of sulfide weathering whereas mechanical disturbance had a major impact on the mobility of the oxide forming elements As, Mo, Sb, U and V. Additionally, options were demonstrated to address with the system the size fractionation of metal(loid)s in pore water samples and the speciation of As(III/V) and Sb(III/V). In the second part, the focus, with a similar experimental design, was placed on the processes leading to the release of metal(loid)s. For this purpose, two incubation experiments with different oxygen supply were conducted in parallel. For the first time the nonmetals carbon, phosphorus and sulfur were analyzed simultaneous to 13 metal(loid)s in sediment pore water by ICP-QQQ-MS. Throughout the experiment metal(loid) size fractionation was monitored. It was confirmed that resuspension promotes the mobility of metalloids such as As, Sb and V, while the release of most metals was largely attributed to pyrite weathering. The colloidal (0.45-16 μm) contribution in terms of mobilization was only relevant for a few elements. Finally, the sampling system was used as part of a new approach to sediment risk assessment. Undisturbed sediment cores from differently contaminated positions in the Trave estuary were examined, considering 16 metal(loid)s, the non-metals C, P and S and the ions NH4+, PO43- and SO42-. By the first in-depth comparison with in-situ dialysis-based pore water sampling the ability of the suction-based approach to represent field conditions was proven. The pore water studies together with supplementing resuspension experiments in bio-geochemical microcosms and sequential extraction identified the most “pristine” sediment of the study area as posing the greatest risk of metal(loid) release. However, the potentially released amounts per kg of sediment are only a few parts per thousand of the average daily loads of the Trave river.

The flexible integration of information from distributed and complex information systems poses a major challenge for organisations. The ontology-based information integration concept SoNBO (Social Network of Business Objects) developed and presented in this dissertation addresses these challenges. In an ontology-based concept, the data structure in the source systems (e.g. operational application systems) is described with the help of a schema (= ontology). The ontology and the data from the source systems can be used to create a (virtualised or materialised) knowledge graph, which is used for information access. The schema can be flexibly adapted to the changing needs of a company regarding their information integration. SoNBO differs from existing concepts known from the Semantic Web (OBDA = Ontology-based Data Access, EKG = Enterprise Knowledge Graph) both in the structure of the company-specific ontology (= Social Network of Concepts) as well as in the structure of the user-specific knowledge graph (= Social Network of Business Objects) and makes use of social principles (known from Enterprise Social Software). Following a Design Science Research approach, the SoNBO framework was developed and the findings documented in this dissertation. The framework provides guidance for the introduction of SoNBO in a company and the knowledge gained from the evaluation (in the company KOSMOS Verlag) is used to demonstrate its viability. The results (SoNBO concept and SoNBO framework) are based on the synthesis of the findings from a structured literature review and the investigation of the status quo of ontology-based information integration in practice: For the status quo in practice, the basic idea of SoNBO is demonstrated in an in-depth case study about the engineering office Vössing, which has been using a self-developed SoNBO application for a few years. The status quo in the academic literature is presented in the form of a structured literature analysis on ontology-based information integration approaches. This dissertation adds to theory in the field of ontology-based information integration approaches (e. g. by an evaluated artefact) and provides an evaluated artefact (the SoNBO Framework) for practice.