A partir da publicação da Crítica da Razão Pura, Kant, na tentativa de colocar a metafísica no caminho da ciência e determinar o que podemos conhecer, recoloca o problema do conhecimento sob uma distinção fundamental: como as coisas são em si mesmas e como essas coisas aparecem para nós. Kant defende, no idealismo transcendental, que nós só podemos conhecer as coisas como elas aparecem, e não podemos conhecê-las como são. Segundo ele, temos uma estrutura a priori determinada, que possibilita o conhecimento das coisas de certa maneira, a saber, sob essas nossas condições de experimentá-las e que, ao mesmo tempo, impossibilita-nos sairmos do nosso ponto de vista e conhecer as coisas como são nelas mesmas, isto é, abstraindo essas condições – e, porque o modo que as coisas aparecem para nós depende dessas condições, se abstrairmos tais condições, as coisas apareceriam de outra maneira; maneira a qual, portanto, não podemos conhecer. Assim, só podemos fazer ciência das coisas como aparecem, e não como são nelas mesmas. Para Hegel, se partimos desse pressuposto, tudo o que podemos conseguir produzir são meras opiniões, e não ciência: é preciso que possamos saber como as coisas são nelas mesmas para que haja conhecimento. Segundo Hegel, podemos conhecer as coisas como elas são, não apenas porque temos condições subjetivas de conhecê-las como elas aparecem, mas porque elas aparecem como são para nós. Sendo assim, o objetivo desta pesquisa é reconstruir a ressignificação hegeliana do aparecer e suas variações conceituais, mais precisamente os conceitos de aparência (Schein) e de fenômeno (Erscheinung) sob a óptica da Ciência da Lógica hegeliana. Para tanto, a presente pesquisa divide-se em três capítulos centrais. No primeiro capítulo, investigaremos o que Kant, na Crítica da Razão Pura, e Hegel, principalmente na Fenomenologia do Espírito, compreendem por ciência e sua relação com a Filosofia. No segundo capítulo, investigaremos o que ambos compreendem por lógica e seu papel para o conhecimento, buscando métodos diferentes para o desenvolvimento de seus sistemas, a partir da Crítica e da Ciência da Lógica. No terceiro capítulo, estabelecemos como Hegel ressignifica o aparecer e o apresenta como um processo lógico de aparecimento dos objetos no mundo, a partir de um aprofundamento da Doutrina da Essência da Ciência da Lógica, buscando, por fim, esclarecer precisamente porque, para Hegel, as distinções kantianas de fenômeno e coisa em si não cumprem o papel de colaborar para a Filosofia seguir o caminho de uma Ciência. / Since the publication of Critique of Pure Reason, Kant, in an attempt to place metaphysics in the way to science and determine what we can know, replace the problem of knowledge in a fundamental distinction: how things are in themselves and how those things appear to us. In transcendental idealism, Kant advocates that we only can know the things as they look like and cannot know them as they are. According to him, we have a determined a priori structure that enables us to know things in certain way, namely, into our conditions to experience them, and that, at the same time, preclude us from leave our point of view and know things as they are, that is, abstracting these conditions – and, for the reason the way things appear to us depends on these conditions, if we abstract them, things would appear in another way: a way that we cannot know. Therefore, we can only make science of things as they appear, and cannot make science of things as they are in themselves. To Hegel, if we start from this presupposition, we can only produce mere opinions, and not science; we need to know how things are in themselves if we want to have knowledge. According to Hegel, we can know things as they are, not only because we have subjective conditions to know them how they appear, but because they appear as they are to us. Thus, the objective of this work is to rebuild the Hegelian reframing of the appear (Scheinen), and its conceptual variations, precisely the concepts of appearance (Schein) and appearance (Erscheinung) in the point of view of the Science of Logic. For this purpose, this work is divided in three central chapters. In the first chapter, we will inquire Kant’s and Hegel’s understanding of science and its relation to Philosophy, mainly in Critic of Pure Reason and Phenomenology of Spirit. In the second chapter, we will inquire their comprehension of logic and its role to knowledge, pursuing different methods to the development of their systems, according to the Critic and Science of Logic. In the third chapter, from a deep reading of the Doctrine of Essence, in the Science of Logic, we will set up how Hegel reframes the appear (Scheinen), and presents it as a logical process of appearance of objects in the world. By the end, we will try to clarify precisely why, to Hegel, the Kantian distinctions of appearance (Erscheinung) and thing in themselves do not play the role of helping Philosophy to follow the way to Science.

A partir da publicação da Crítica da Razão Pura, Kant, na tentativa de colocar a metafísica no caminho da ciência e determinar o que podemos conhecer, recoloca o problema do conhecimento sob uma distinção fundamental: como as coisas são em si mesmas e como essas coisas aparecem para nós. Kant defende, no idealismo transcendental, que nós só podemos conhecer as coisas como elas aparecem, e não podemos conhecê-las como são. Segundo ele, temos uma estrutura a priori determinada, que possibilita o conhecimento das coisas de certa maneira, a saber, sob essas nossas condições de experimentá-las e que, ao mesmo tempo, impossibilita-nos sairmos do nosso ponto de vista e conhecer as coisas como são nelas mesmas, isto é, abstraindo essas condições – e, porque o modo que as coisas aparecem para nós depende dessas condições, se abstrairmos tais condições, as coisas apareceriam de outra maneira; maneira a qual, portanto, não podemos conhecer. Assim, só podemos fazer ciência das coisas como aparecem, e não como são nelas mesmas. Para Hegel, se partimos desse pressuposto, tudo o que podemos conseguir produzir são meras opiniões, e não ciência: é preciso que possamos saber como as coisas são nelas mesmas para que haja conhecimento. Segundo Hegel, podemos conhecer as coisas como elas são, não apenas porque temos condições subjetivas de conhecê-las como elas aparecem, mas porque elas aparecem como são para nós. Sendo assim, o objetivo desta pesquisa é reconstruir a ressignificação hegeliana do aparecer e suas variações conceituais, mais precisamente os conceitos de aparência (Schein) e de fenômeno (Erscheinung) sob a óptica da Ciência da Lógica hegeliana. Para tanto, a presente pesquisa divide-se em três capítulos centrais. No primeiro capítulo, investigaremos o que Kant, na Crítica da Razão Pura, e Hegel, principalmente na Fenomenologia do Espírito, compreendem por ciência e sua relação com a Filosofia. No segundo capítulo, investigaremos o que ambos compreendem por lógica e seu papel para o conhecimento, buscando métodos diferentes para o desenvolvimento de seus sistemas, a partir da Crítica e da Ciência da Lógica. No terceiro capítulo, estabelecemos como Hegel ressignifica o aparecer e o apresenta como um processo lógico de aparecimento dos objetos no mundo, a partir de um aprofundamento da Doutrina da Essência da Ciência da Lógica, buscando, por fim, esclarecer precisamente porque, para Hegel, as distinções kantianas de fenômeno e coisa em si não cumprem o papel de colaborar para a Filosofia seguir o caminho de uma Ciência. / Since the publication of Critique of Pure Reason, Kant, in an attempt to place metaphysics in the way to science and determine what we can know, replace the problem of knowledge in a fundamental distinction: how things are in themselves and how those things appear to us. In transcendental idealism, Kant advocates that we only can know the things as they look like and cannot know them as they are. According to him, we have a determined a priori structure that enables us to know things in certain way, namely, into our conditions to experience them, and that, at the same time, preclude us from leave our point of view and know things as they are, that is, abstracting these conditions – and, for the reason the way things appear to us depends on these conditions, if we abstract them, things would appear in another way: a way that we cannot know. Therefore, we can only make science of things as they appear, and cannot make science of things as they are in themselves. To Hegel, if we start from this presupposition, we can only produce mere opinions, and not science; we need to know how things are in themselves if we want to have knowledge. According to Hegel, we can know things as they are, not only because we have subjective conditions to know them how they appear, but because they appear as they are to us. Thus, the objective of this work is to rebuild the Hegelian reframing of the appear (Scheinen), and its conceptual variations, precisely the concepts of appearance (Schein) and appearance (Erscheinung) in the point of view of the Science of Logic. For this purpose, this work is divided in three central chapters. In the first chapter, we will inquire Kant’s and Hegel’s understanding of science and its relation to Philosophy, mainly in Critic of Pure Reason and Phenomenology of Spirit. In the second chapter, we will inquire their comprehension of logic and its role to knowledge, pursuing different methods to the development of their systems, according to the Critic and Science of Logic. In the third chapter, from a deep reading of the Doctrine of Essence, in the Science of Logic, we will set up how Hegel reframes the appear (Scheinen), and presents it as a logical process of appearance of objects in the world. By the end, we will try to clarify precisely why, to Hegel, the Kantian distinctions of appearance (Erscheinung) and thing in themselves do not play the role of helping Philosophy to follow the way to Science.

A partir da publicação da Crítica da Razão Pura, Kant, na tentativa de colocar a metafísica no caminho da ciência e determinar o que podemos conhecer, recoloca o problema do conhecimento sob uma distinção fundamental: como as coisas são em si mesmas e como essas coisas aparecem para nós. Kant defende, no idealismo transcendental, que nós só podemos conhecer as coisas como elas aparecem, e não podemos conhecê-las como são. Segundo ele, temos uma estrutura a priori determinada, que possibilita o conhecimento das coisas de certa maneira, a saber, sob essas nossas condições de experimentá-las e que, ao mesmo tempo, impossibilita-nos sairmos do nosso ponto de vista e conhecer as coisas como são nelas mesmas, isto é, abstraindo essas condições – e, porque o modo que as coisas aparecem para nós depende dessas condições, se abstrairmos tais condições, as coisas apareceriam de outra maneira; maneira a qual, portanto, não podemos conhecer. Assim, só podemos fazer ciência das coisas como aparecem, e não como são nelas mesmas. Para Hegel, se partimos desse pressuposto, tudo o que podemos conseguir produzir são meras opiniões, e não ciência: é preciso que possamos saber como as coisas são nelas mesmas para que haja conhecimento. Segundo Hegel, podemos conhecer as coisas como elas são, não apenas porque temos condições subjetivas de conhecê-las como elas aparecem, mas porque elas aparecem como são para nós. Sendo assim, o objetivo desta pesquisa é reconstruir a ressignificação hegeliana do aparecer e suas variações conceituais, mais precisamente os conceitos de aparência (Schein) e de fenômeno (Erscheinung) sob a óptica da Ciência da Lógica hegeliana. Para tanto, a presente pesquisa divide-se em três capítulos centrais. No primeiro capítulo, investigaremos o que Kant, na Crítica da Razão Pura, e Hegel, principalmente na Fenomenologia do Espírito, compreendem por ciência e sua relação com a Filosofia. No segundo capítulo, investigaremos o que ambos compreendem por lógica e seu papel para o conhecimento, buscando métodos diferentes para o desenvolvimento de seus sistemas, a partir da Crítica e da Ciência da Lógica. No terceiro capítulo, estabelecemos como Hegel ressignifica o aparecer e o apresenta como um processo lógico de aparecimento dos objetos no mundo, a partir de um aprofundamento da Doutrina da Essência da Ciência da Lógica, buscando, por fim, esclarecer precisamente porque, para Hegel, as distinções kantianas de fenômeno e coisa em si não cumprem o papel de colaborar para a Filosofia seguir o caminho de uma Ciência. / Since the publication of Critique of Pure Reason, Kant, in an attempt to place metaphysics in the way to science and determine what we can know, replace the problem of knowledge in a fundamental distinction: how things are in themselves and how those things appear to us. In transcendental idealism, Kant advocates that we only can know the things as they look like and cannot know them as they are. According to him, we have a determined a priori structure that enables us to know things in certain way, namely, into our conditions to experience them, and that, at the same time, preclude us from leave our point of view and know things as they are, that is, abstracting these conditions – and, for the reason the way things appear to us depends on these conditions, if we abstract them, things would appear in another way: a way that we cannot know. Therefore, we can only make science of things as they appear, and cannot make science of things as they are in themselves. To Hegel, if we start from this presupposition, we can only produce mere opinions, and not science; we need to know how things are in themselves if we want to have knowledge. According to Hegel, we can know things as they are, not only because we have subjective conditions to know them how they appear, but because they appear as they are to us. Thus, the objective of this work is to rebuild the Hegelian reframing of the appear (Scheinen), and its conceptual variations, precisely the concepts of appearance (Schein) and appearance (Erscheinung) in the point of view of the Science of Logic. For this purpose, this work is divided in three central chapters. In the first chapter, we will inquire Kant’s and Hegel’s understanding of science and its relation to Philosophy, mainly in Critic of Pure Reason and Phenomenology of Spirit. In the second chapter, we will inquire their comprehension of logic and its role to knowledge, pursuing different methods to the development of their systems, according to the Critic and Science of Logic. In the third chapter, from a deep reading of the Doctrine of Essence, in the Science of Logic, we will set up how Hegel reframes the appear (Scheinen), and presents it as a logical process of appearance of objects in the world. By the end, we will try to clarify precisely why, to Hegel, the Kantian distinctions of appearance (Erscheinung) and thing in themselves do not play the role of helping Philosophy to follow the way to Science.

The field of self-contained linear hydraulic drives based on variable-speed electrical motors and fixed displacement pumps is gaining interest from both industry and academia. Some of the main reasons for this is the possibility to improve the energy efficiency of such drives compared to conventional valve controlled drives, and the possibility for electrical regeneration allowing power sharing between multiple drives [1]. The main drawback for such types of drive concepts is a low pressure in the nonload carrying cylinder chamber. This induces a low drive stiffness limiting the achievable drive bandwidth and hence the application range. However, a so-called self-locking compact drive architecture recently proposed allows maintaining a proper drive stiffness by virtue of separate forward and return flow paths, combining the advantages of efficient flow control into the cylinder and a throttle driven flow out of the cylinder. The multiple inputs available in this architecture allow the control to target several objectives concurrently, for example piston motion, drive stiffness and fluid temperature. The purpose of the study presented is to analyse the dynamic couplings between the control objectives via relative gain array (RGA) methods, and the realization of input- and output transformations effectively decoupling relevant dynamic interactions. These transformations allow the usage of simple SISO-controllers for each control objective, and a method for controlling motion and fluid temperature concurrently, is proposed and experimentally verified.

This paper presents a conceptual study aiming to improve the compactness of electro-hydraulic compact drives (ECD ). In most current ECD architectures, gas accumulators are used as volume compensators for the flow imbalance emerging whenever asymmetric single rod cylinders are used. To stay within a required reservoir pressure range typically from two to four bar, a large gas volume is required, compromising system compactness. Combining conventional ECD architectures with a bootstrap reservoir offers a greater degree of freedom in system design, which enables downsizing or avoidance of the gas volume. Another potential benefit by including a bootstrap reservoir is the possibility of elevating the backpressure of the ECD thus enhancing drive stiffness, expanding the application range and market acceptance. Based on an open analysis of the solution space occurring when introducing a bootstrap reservoir, three system architectures are selected for a conceptual study. The results show that the downsizing potential is strongly dependent on the maximum friction force and the area ratio of the bootstrap reservoir pistons, while a linear analysis reveals that for some system architectures the bootstrap reservoir may severely influence the system dynamics. Simulation results confirm the functionality of the proposed system architectures, and show that a potential for downsizing/avoiding the gas volume, as well as increasing the ECD stiffness is present.

Die Bachelorarbeit behandelt die Themen der Selbstorganisation in Produktionssystemen im Kontext von Industrie 4.0. Dabei wird gezeigt, wie man mithilfe von einer Ameisenkolonieoptimierung die Reihenfolgeplanung organisieren kann.:Abbildungsverzeichnis Tabellenverzeichnis Formelverzeichnis 1. Einleitung 1.1. Motivation 1.2. Ziele 1.3. Vorgehensweise 2. Sich selbst organisierende Produktionen 2.1. Begriffserklärung 2.2. Stand der Technik 2.3. Reihenfolgeplanung als ein Problem der Selbstorganisation 2.3.1. Begriffserklärung 2.3.2. Stand der Technik 2.3.3. Umsetzung in einer Selbstorganisation 3. Ameisenkolonieoptimierung 3.1. Begriffserklärung 3.2. Allgemeine Umsetzung 3.3. Konkrete Umsetzungen 3.4. Vor- und Nachteile 3.5. Anwendungsbeispiele 4. Entwicklung einer Ameisenkolonieoptimierung für ein sich selbst organisierendes Produktionssystem 4.1. Analyse des gegebenen sich selbst organisierenden Produktionssystems 4.1.1. Grobanalyse des Systems 4.1.2. Feinanalyse der bisherigen Reihenfolgeplanung 4.2. Entwurf der Reihenfolgeplanung durch Prinzipien der Ameisenkolonieoptimierung 4.3. Implementierung der Prinzipien der Ameisenkolonieoptimierung 5. Empirische Untersuchung der implementierten Ameisenkolonieoptimierung 5.1. Beschreibung der gegebenen Produktionsdaten 5.2. Szenarienuntersuchung zur Funktionsfähigkeit 5.2.1. Schichtwechselszenario 5.2.2. Abnutzungs- und Wartungsszenario 5.2.3. Vergleichsszenario 5.3. Untersuchung hinsichtlich der Laufzeit und des Speicherbedarfs 5.3.1. Laufzeit 5.3.2. Speicherbedarf 6. Zusammenfassung und Ausblick 6.1. Zusammenfassung 6.2. Ausblick Quellenverzeichnis / The bachelor thesis is about self organization in production systems in the context of Industry 4.0. Its about ant colony optimization for scheduling in the production planning.:Abbildungsverzeichnis Tabellenverzeichnis Formelverzeichnis 1. Einleitung 1.1. Motivation 1.2. Ziele 1.3. Vorgehensweise 2. Sich selbst organisierende Produktionen 2.1. Begriffserklärung 2.2. Stand der Technik 2.3. Reihenfolgeplanung als ein Problem der Selbstorganisation 2.3.1. Begriffserklärung 2.3.2. Stand der Technik 2.3.3. Umsetzung in einer Selbstorganisation 3. Ameisenkolonieoptimierung 3.1. Begriffserklärung 3.2. Allgemeine Umsetzung 3.3. Konkrete Umsetzungen 3.4. Vor- und Nachteile 3.5. Anwendungsbeispiele 4. Entwicklung einer Ameisenkolonieoptimierung für ein sich selbst organisierendes Produktionssystem 4.1. Analyse des gegebenen sich selbst organisierenden Produktionssystems 4.1.1. Grobanalyse des Systems 4.1.2. Feinanalyse der bisherigen Reihenfolgeplanung 4.2. Entwurf der Reihenfolgeplanung durch Prinzipien der Ameisenkolonieoptimierung 4.3. Implementierung der Prinzipien der Ameisenkolonieoptimierung 5. Empirische Untersuchung der implementierten Ameisenkolonieoptimierung 5.1. Beschreibung der gegebenen Produktionsdaten 5.2. Szenarienuntersuchung zur Funktionsfähigkeit 5.2.1. Schichtwechselszenario 5.2.2. Abnutzungs- und Wartungsszenario 5.2.3. Vergleichsszenario 5.3. Untersuchung hinsichtlich der Laufzeit und des Speicherbedarfs 5.3.1. Laufzeit 5.3.2. Speicherbedarf 6. Zusammenfassung und Ausblick 6.1. Zusammenfassung 6.2. Ausblick Quellenverzeichnis

Superpixels are the results of an image oversegmentation. They are an established intermediate level image representation and used for various applications including object detection, 3d reconstruction and semantic segmentation. While there are various approaches to create such segmentations, there is a lack of knowledge about their properties. In particular, there are contradicting results published in the literature. This thesis identifies segmentation quality, stability, compactness and runtime to be important properties of superpixel segmentation algorithms. While for some of these properties there are established evaluation methodologies available, this is not the case for segmentation stability and compactness. Therefore, this thesis presents two novel metrics for their evaluation based on ground truth optical flow. These two metrics are used together with other novel and existing measures to create a standardized benchmark for superpixel algorithms. This benchmark is used for extensive comparison of available algorithms. The evaluation results motivate two novel segmentation algorithms that better balance trade-offs of existing algorithms: The proposed Preemptive SLIC algorithm incorporates a local preemption criterion in the established SLIC algorithm and saves about 80 % of the runtime. The proposed Compact Watershed algorithm combines Seeded Watershed segmentation with compactness constraints to create regularly shaped, compact superpixels at the even higher speed of the plain watershed transformation. Operating autonomous systems over the course of days, weeks or months, based on visual navigation, requires repeated recognition of places despite severe appearance changes as they are for example induced by illumination changes, day-night cycles, changing weather or seasons - a severe problem for existing methods. Therefore, the second part of this thesis presents two novel approaches that incorporate superpixel segmentations in place recognition in changing environments. The first novel approach is the learning of systematic appearance changes. Instead of matching images between, for example, summer and winter directly, an additional prediction step is proposed. Based on superpixel vocabularies, a predicted image is generated that shows, how the summer scene could look like in winter or vice versa. The presented results show that, if certain assumptions on the appearance changes and the available training data are met, existing holistic place recognition approaches can benefit from this additional prediction step. Holistic approaches to place recognition are known to fail in presence of viewpoint changes. Therefore, this thesis presents a new place recognition system based on local landmarks and Star-Hough. Star-Hough is a novel approach to incorporate the spatial arrangement of local image features in the computation of image similarities. It is based on star graph models and Hough voting and particularly suited for local features with low spatial precision and high outlier rates as they are expected in the presence of appearance changes. The novel landmarks are a combination of local region detectors and descriptors based on convolutional neural networks. This thesis presents and evaluates several new approaches to incorporate superpixel segmentations in local region detection. While the proposed system can be used with different types of local regions, in particular the combination with regions obtained from the novel multiscale superpixel grid shows to perform superior to the state of the art methods - a promising basis for practical applications.

Interleukin-10 (IL-10) is an important regulator of immune responses secreted by different cell types. Previous results from our group suggested that the biological effects of this cytokine critically depend on its cellular source. Recent studies reported IL-10 dependent immunosuppressive functions of a specialized subset of regulatory B cells and mast cells. These results relied on adoptive cell transfers, a technique which can potentially introduce artifacts. Therefore, we aimed to readdress these questions in independent models using IL-10 transcriptional reporter mice and various conditional IL-10 mutant mice. Findings in IL-10 reporter system suggested prominent IL-10 transcription in regulatory B cells upon LPS administration. Exposure of mice to contact allergen revealed robust reporter expression in CD8 T cells, moderate to mild reporter expression in CD4 T cells and dendritic cells (DC) respectively, and lack of reporter expression in B cells, mast cells and NK cells in allergen challenged ears. We generated cell-type specific IL-10 mutants by Cre/LoxP-mediated conditional gene inactivation. Efficiency and specificity of Cre-mediated recombination was demonstrated by Southern blot and PCR methods. Various immunogenic challenges in conditional IL-10 mutants did not reveal a role for B cell-derived IL-10 in restraining innate TLR or T cell-dependent inflammatory responses. Likewise, mice with selective inactivation of the il10 gene in mast cells exhibited normal CHS responses and unaltered immune response to CpG oligodeoxynucleotides. On the other hand, DC-specific IL-10 mutants developed excessive inflammatory responses to contact allergens, while innate responses to TLR ligands were not altered. This indicates a non-redundant role for DC-derived IL-10 in contact allergy. Thus, the conditional IL-10 ‘‘knockout’’ mice combined with the novel transcriptional IL-10 reporter system can serve as ideal tools to understand the cell-type specific contributions to IL-10-mediated immune regulation.

Superpixels are the results of an image oversegmentation. They are an established intermediate level image representation and used for various applications including object detection, 3d reconstruction and semantic segmentation. While there are various approaches to create such segmentations, there is a lack of knowledge about their properties. In particular, there are contradicting results published in the literature. This thesis identifies segmentation quality, stability, compactness and runtime to be important properties of superpixel segmentation algorithms. While for some of these properties there are established evaluation methodologies available, this is not the case for segmentation stability and compactness. Therefore, this thesis presents two novel metrics for their evaluation based on ground truth optical flow. These two metrics are used together with other novel and existing measures to create a standardized benchmark for superpixel algorithms. This benchmark is used for extensive comparison of available algorithms. The evaluation results motivate two novel segmentation algorithms that better balance trade-offs of existing algorithms: The proposed Preemptive SLIC algorithm incorporates a local preemption criterion in the established SLIC algorithm and saves about 80 % of the runtime. The proposed Compact Watershed algorithm combines Seeded Watershed segmentation with compactness constraints to create regularly shaped, compact superpixels at the even higher speed of the plain watershed transformation. Operating autonomous systems over the course of days, weeks or months, based on visual navigation, requires repeated recognition of places despite severe appearance changes as they are for example induced by illumination changes, day-night cycles, changing weather or seasons - a severe problem for existing methods. Therefore, the second part of this thesis presents two novel approaches that incorporate superpixel segmentations in place recognition in changing environments. The first novel approach is the learning of systematic appearance changes. Instead of matching images between, for example, summer and winter directly, an additional prediction step is proposed. Based on superpixel vocabularies, a predicted image is generated that shows, how the summer scene could look like in winter or vice versa. The presented results show that, if certain assumptions on the appearance changes and the available training data are met, existing holistic place recognition approaches can benefit from this additional prediction step. Holistic approaches to place recognition are known to fail in presence of viewpoint changes. Therefore, this thesis presents a new place recognition system based on local landmarks and Star-Hough. Star-Hough is a novel approach to incorporate the spatial arrangement of local image features in the computation of image similarities. It is based on star graph models and Hough voting and particularly suited for local features with low spatial precision and high outlier rates as they are expected in the presence of appearance changes. The novel landmarks are a combination of local region detectors and descriptors based on convolutional neural networks. This thesis presents and evaluates several new approaches to incorporate superpixel segmentations in local region detection. While the proposed system can be used with different types of local regions, in particular the combination with regions obtained from the novel multiscale superpixel grid shows to perform superior to the state of the art methods - a promising basis for practical applications.

Visual place recognition is the task of finding same places in a set of database images for a given set of query images. This becomes particularly challenging for long-term applications when the environmental condition changes between or within the database and query set, e.g., from day to night. Visual place recognition in changing environments can be used if global position data like GPS is not available or very inaccurate, or for redundancy. It is required for tasks like loop closure detection in SLAM, candidate selection for global localization, or multi-robot/multi-session mapping and map merging. In contrast to pure image retrieval, visual place recognition can often build upon additional information and data for improvements in performance, runtime, or memory usage. This includes additional data-inherent knowledge about information that is contained in the image sets themselves because of the way they were recorded. Using data-inherent knowledge avoids the dependency on other sensors, which increases the generality of methods for an integration into many existing place recognition pipelines. This thesis focuses on the usage of additional data-inherent knowledge. After the discussion of basics about visual place recognition, the thesis gives a systematic overview of existing data-inherent knowledge and corresponding methods. Subsequently, the thesis concentrates on a deeper consideration and exploitation of four different types of additional data-inherent knowledge. This includes 1) sequences, i.e., the database and query set are recorded as spatio-temporal sequences so that consecutive images are also adjacent in the world, 2) knowledge of whether the environmental conditions within the database and query set are constant or continuously changing, 3) intra-database similarities between the database images, and 4) intra-query similarities between the query images. Except for sequences, all types have received only little attention in the literature so far. For the exploitation of knowledge about constant conditions within the database and query set (e.g., database: summer, query: winter), the thesis evaluates different descriptor standardization techniques. For the alternative scenario of continuous condition changes (e.g., database: sunny to rainy, query: sunny to cloudy), the thesis first investigates the qualitative and quantitative impact on the performance of image descriptors. It then proposes and evaluates four unsupervised learning methods, including our novel clustering-based descriptor standardization method K-STD and three PCA-based methods from the literature. To address the high computational effort of descriptor comparisons during place recognition, our novel method EPR for efficient place recognition is proposed. Given a query descriptor, EPR uses sequence information and intra-database similarities to identify nearly all matching descriptors in the database. For a structured combination of several sources of additional knowledge in a single graph, the thesis presents our novel graphical framework for place recognition. After the minimization of the graph's error with our proposed ICM-based optimization, the place recognition performance can be significantly improved. For an extensive experimental evaluation of all methods in this thesis and beyond, a benchmark for visual place recognition in changing environments is presented, which is composed of six datasets with thirty sequence combinations.