@mastersthesis { Ama2018a,
     author        = { Sabarish Kumar Amaravadi },
     title         = { Enhancement of Real-Time Object Tracking Using Local Image Features },
     month         = { April },
     year          = { 2019 },
     school        = { Paderborn University },
     type          = { Master's thesis },
     abstract      = { The aim of this master’s thesis is to improve the accuracy of an object tracking algorithm by integrating (additional) local image features in a modular fashion. The object tracking algorithm should have the following specific properties: use of local image features, use of a single template for initialization, use of structural information, and real-time capability. The first and foremost step is to identify an object tracking algorithm with these properties for which an implementation exists. The second step is to integrate additional local image features into the algorithm, for instance, using parallel integration. Finally, the resulting object tracking algorithm will be tested with a range of standard data sets and is compared with its original version and other state-of-the-art object tracking algorithms. The resulting object tracking algorithm should effectively deal with challenges such as occlusion, clutter, and the disappearance of objects of interest. }
}

@misc { BIPM19,
     author        = { Bureau International des Poids et Mesures BIPM },
     title         = { <a href="https://www.bipm.org/en/publications/si-brochure/">The International System of Units (SI)</a> },
     month         = { January },
     year          = { 2019 }
}

@mastersthesis { Biswas18,
     author        = { Hridkamol Biswas },
     title         = { Learning Condition-invariant Scene Representations for Across the Seasons Place Recognition },
     month         = { January },
     year          = { 2019 },
     school        = { Paderborn University },
     type          = { Master's thesis },
     abstract      = { For autonomous navigation in an unknown environment, a mobile robot has to build a map of the environments and localize itself within the map. When a robot movies in an environment, the errors in motion accumulate over time, which gradually makes the map inconsistent. To reduce the error in estimating the map of the environment, it is necessary for a robot to recognize previously visited locations, called as a loop-closure detection. When an environment goes through extreme perceptual changes, e.g., seasonal changes, it becomes a harder problem for a mobile robot to recognize its previously visited locations. The purpose of this research work is to develop an algorithm that learns the robust representation of the scenes in presence of the seasonal changes to assist the correct loop-closure detection. To reach this goal, independent component analysis and autoencoder have been adopted. In this work, it has been observed that, the ICA has the promising capabilities of extracting the condition-invariant image descriptors when a robot is operating across the wide range of appearance changing environment. The potential of distinguishing the condition-variant and condition-invariant features of a scene by an autoencoder has been investigated, which shows a considerable performance and discovers a possible route to take this work to the next step. The performance is evaluated with the baseline method using the precision-recall curves and fraction of correct matches. For the selected evaluation criteria, our algorithm showed superior performance compared to the baseline method. }
}

@mastersthesis { Bri2018a,
     author        = { Lukas Brinkmann },
     title         = { Vorverarbeitung nat{"u}rlicher Kantenbilder zur echtzeitf{"a}higen Objekterkennung },
     month         = { February },
     year          = { 2019 },
     school        = { Paderborn University },
     abstract      = { Im Rahmen dieser Studienarbeit soll ein bestehendes Verfahren zur Vorverarbeitung nat{"u}rlicher Kantenbilder zur echtzeitf{"a}higen Analyse von Objektkonturen angepasst und erweitert werden. Im GET Lab wurde bereits ein rekursives Verfahren umgesetzt, mit dem die Kanten in bin{"a}ren Kantenbildern gezielt eingelesen und anschlie{\ss}end weiter analysiert werden k{"o}nnen. Bei nat{"u}rlichen Kantenbildern tritt jedoch h{"a}ufig das Problem auf, dass die Kanten Unterbrechungen aufweisen und dadurch eine anschlie{\ss}ende Konturanalyse erschweren. Zudem k{"o}nnen Kantenkreuzungen und Verzweigungen auftreten, so dass der weitere Verlauf einer Kante nicht bekannt oder mehrdeutig ist. Bei der {"U}berarbeitung des bestehenden Verfahrens k{"o}nnen einige Klassen und Funktionen {"u}bernommen werden. Die bisherige Umsetzung des rekursiven Verfahrens ist jedoch nur eingeschr{"a}nkt f{"u}r weitere Verarbeitungsschritte geeignet, da sich unter anderem aufeinander zulaufende Rekursionen mit der aktuellen Implementierung nicht verarbeiten lassen. }
}

@mastersthesis { DAU2018,
     author        = { Christian Daube },
     title         = { Semantic Motion Segmentation using Deep Learning },
     month         = { January },
     year          = { 2019 },
     school        = { Paderborn University },
     type          = { Master's thesis },
     abstract      = { The analysis of complex and challenging environments by autonomous systems requires fast and accurate algorithms for the detection of semantic classes and motion states. In previous works, conventional methods used constraints to perform semantic motion segmentation. These constraints limit the results of the motion segmentation approaches to specific scenarios. As a result, more general approaches without limiting constraints are required to handle more challenging and variable environments. Recently, deep convolutional neural networks have been used for object detection in autonomous systems acting in dynamic real-world environments. These neural networks provide a superior performance in comparison to previous approaches. In this thesis, the task is to design a deep convolutional neural network, which incorporates optical flow and semantic segmentation to detect moving objects. Thus, the neural network learns to predict the semantic class as well as the motion state at each pixel from a pair of consecutive images. Finally, the proposed neural network is evaluated and examined with labeled datasets. Furthermore, it is tested on a real-world environment using a camera mounted on a car. The results show real-time performance for the semantic motion segmentation, while providing reasonably good performance. }
}

@book { Flo19,
     author        = { Thomas L. Floyd and David M. Buchla },
     title         = { Principles of Electric Circuits - Conventional Current Version },
     month         = { January },
     year          = { 2019 },
     edition       = { 10. Auflage },
     publisher     = { Pearson Higher Education USA },
     isbn          = { 978-0134879482 }
}

@mastersthesis { Foer2018,
     author        = { Christopher F{"o}rster },
     title         = { Entwicklung und Vergleich von Methoden zur autonomen Navigation in unwegsamem Gel{"a}nde f{"u}r einen kettengetriebenen Roboter unter besonderer Ber{"u}cksichtigung maschinellen Lernens },
     month         = { November },
     year          = { 2019 },
     school        = { Paderborn University },
     type          = { Master's thesis },
     abstract      = { F{"u}r Such- und Bergungsaufgaben werden Roboter ben{"o}tigt, die in der Lage sind, komplexe Aufgaben selbstst{"a}ndig zu erf{"u}llen. Eine besondere Herausforderung ist dabei die Navigation. Deshalb ist in dieser Arbeit ein komplett autonomes System entstanden, welches es Robotern erm{"o}glichen soll, in unwegsamem Gel{"a}nde zu navigieren. Hierzu wurden zwei grundlegend verschiedene Ans{"a}tze verwendet. Ein Ansatz nutzt herk{"o}mmliche Verfahren, um mit einer H{"o}henkarte Befahrbarkeitseigenschaften zu berechnen, daraus einen Pfad zu planen und diesem dann zu verfolgen. Der andere Ansatz nutzt Deep Reinforcement Learning, um auf Basis komplexer, in einer Simulationsumgebung bereitgestellter Sensorinformationen, wie H{"o}henkarte und Tiefenbilder, ein autonomes Navigationsverhalten zu erlernen. Weiterhin wurde f{"u}r beide Systeme eine Steuerung f{"u}r kettenbespannte Hilfsarme entworfen, die es dem Roboter erm{"o}glicht, Hindernisse zu {"u}berwinden. Die Ergebnisse wurden einander abschlie{\ss}end gegen{"u}bergestellt und bewertet.  }
}

@article { kazmi19a,
     author        = { S. M. Ali Musa Kazmi and B{"a}rbel Mertsching },
     title         = { Detecting the Expectancy of a Place using Nearby Context for Appearance-based Mapping },
     month         = { June },
     year          = { 2019 },
     journal       = { IEEE Transactions on Robotics (T-RO) },
     number        = { 6 },
     pages         = { 1352 -- 1366 },
     volume        = { 35 },
     issn          = { 1941-0468 },
     abstract      = { In recent years, place recognition techniques have been extensively studied in the domain of robotic mapping, referred to as appearance-based mapping. Nonetheless, the majority of these methods focus the challenges of place recognition in offline or supervised scenarios, which in certain conditions, e.g., unknown environments, is infeasible. In this research, we address
the challenges of online place recognition and demonstrate the general applicability of our approach in versatile environments. To this end, a modified growing self-organizing network of
neurons is proposed, which incrementally adapts itself to learn the topology of the perceptual space formed by gist features. Given a query image and the network state at any time instant,
the expected activity of the network is estimated using a proposed Bayesian framework, and the current place is categorized as familiar or novel. Exhaustive experiments on eleven challenging
sequences signify the strength of our algorithm for a reliable and real-time place recognition on routes as large as 18 km. Compared to many state-of-the-art approaches, our method does not need offline training or environment-specific parameter tuning. }
}

@inproceedings { kazmi19b,
     author        = { S. M. Ali Musa Kazmi and Mahmoud Mohamed and B{"a}rbel Mertsching },
     title         = { Feature-agnostic Low-cost Place Recognition for Appearance-based Mapping },
     month         = { July },
     year          = { 2019 },
     booktitle     = { 12th International Conference on Computer Vision Systems (ICVS) },
     abstract      = { The agent’s ability to locate itself in an unfamiliar environment is essential for a reliable navigation. To address this challenge, place recognition methods are widely adopted. A common trend among most of these methods is that they are either tailored to work in specific environments or need prior training overhead. Whereas, others demand extreme computational resources, such as CNNs. In this paper, we analyze the existing GSOM-based place recognition framework and investigate the question of translating the system to other feature spaces, such as HOG, for low-cost place recognition. The experiments performed on four challenging sequences demonstrate the algorithm’s ability to learn the representation of the new feature space without parameter tuning, provided the scaling factor along each dimension of the descriptor is taken into account. This highlights the feature-agnostic characteristic of the algorithm. We further observed that despite the low dimensionality of the HOG descriptor, the algorithm shows comparable place recognition results to the gist features, while offering threefold speed-ups in execution time.
 }
}

@mastersthesis { Lu2018,
     author        = { Ke Lu },
     title         = { Gesture-based Control System for a Robot Arm },
     month         = { May },
     year          = { 2019 },
     school        = { Paderborn University },
     type          = { Master's thesis },
     abstract      = { Nowadays, robot arms are being used extensively to execute many precise or dangerous tasks and they are usually controlled by a joystick. However, the traditional control method is complicated and inefficient. This thesis addresses this problem by proposing a new control method for robot arms using hand gestures recognition. The aim is to reduce the difficulty of controlling a robot arm as well as save the learning time for robot arm operators. In this thesis, an RGB-D sensor will be used for acquiring images and obtaining depth information. Using an algorithm of detecting fingertips and palm center, the trajectory of hand movement is then extracted. After the calculations of inverse kinematics, the end-effector of the robot arm follows the same trajectory. Besides, the gripper of the robot arm is controlled by the distance between the fingertips of index finger and thumb. The system performance will be evaluated using the robot arm of GETjag in terms of accuracy and efficiency.
 }
}

@inproceedings { Mahmoud19,
     author        = { Mahmoud Mohamed and B{"a}rbel Mertsching },
     title         = { Robust Optical Flow Estimation Using the Monocular Epipolar Geometry },
     month         = { July },
     year          = { 2019 },
     booktitle     = { 12th International Conference on Computer Vision Systems (ICVS) },
     abstract      = { The estimation of optical flow in cases of illumination change, sparsely-textured regions or fast moving objects is a challenging problem. In this paper, we analyze the use of a texture constancy constraint based on local descriptors (i.e., HOG) integrated with the monocular epipolar geometry to estimate robustly optical flow. The framework is implemented in differential data fidelities using a total variation model in a multi-resolution scheme. Besides, we propose an effective method to refine the fundamental matrix along with the estimation of the optical flow. Experimental results based on the challenging KITTI dataset show that the integration of texture constancy constraint with the monocular epipolar line constraint and the enhancement of the fundamental matrix significantly increases the accuracy of the estimated optical flow. Furthermore, a comparison with existing state-of-the-art approaches shows better performance for the proposed approach. }
}

@phdthesis { MM2019,
     author        = { Mahmoud Mohamed },
     title         = { Robust Motion Estimation for Qualitative Dynamic Scene Analysis },
     month         = { July },
     year          = { 2019 },
     school        = { Paderborn University },
     type          = { Dissertation (PhD) },
     abstract      = { Dynamic scene analysis is the primary challenge for various applications such as Advanced Driver Assistance Systems (ADAS), and in any autonomous robot operation in dynamic environments. Autonomous robot/vehicle can carry out desired tasks without continuous human interaction. Distinctly, robust detection, tracking, and recognition of moving objects as well as an estimation of camera ego-motion in a scene are necessary expendables for many autonomous tasks. For instance, in mobile robotics, moving objects are possibly more insecure than stationary objects for safe navigation. In particular, rescue robot systems could increase their performance enormously if they were capable of interacting with moving victims. Robust detection/tracking of moving objects from a moving camera in an outdoor environment is a challenging task due to dynamically changing cluttered backgrounds, large motion, varying lighting conditions, less texture objects, partial object occlusion, and varying object viewpoints.

The work presented in this thesis cops with the problem of robust estimation of 2D motion and tracking of moving objects with the problems mentioned above. Therefore, this work introduces a new approach to improve the accuracy of the 2D motion estimation, which called optical flow, in case of large motion using the coarse-to-fine technique. The proposed algorithm estimates the optical flow of fast as well as slow objects correctly and with less processing cost. Moreover, the presented work proposes a novel optimization model for the optical flow estimation base on the texture constraint. The texture constraint assumes that object textures such as edges, gradients, or orientation-of-image features remain constant in case of objects or camera motion. The optimization model uses an objective function to minimize dissimilarity between image texture using local descriptors. The proposed model is not limited to any local texture descriptors, for instance, the histogram of oriented gradient (HOG), the modified local directional pattern (MLDP), the census transform, and other descriptors are used. Furthermore, we present the usage of the monocular epipolar line constraint to improve the accuracy of the optical flow in the case of texture-less regions. The new model estimates the optical flow correctly in most cases when most state-of-the-art approaches that depend on the brightness constancy of a pixel fail. Besides, we propose a new approach for detecting and tracking all moving objects. The proposed algorithm works with static as well as a moving camera, and the results show the successful detection, estimation, and tracking of moving objects in indoor and outdoor environments. Several experiments and applications have been conducted to test and evaluate the algorithms extensively. The results have shown that the proposed algorithms outperformed the state-of-the-art approaches based on the standard benchmark datasets.  }
}

@book { PP19,
     author        = { Steffen Paul and Reinhold Paul },
     title         = { Grundlagen der Elektrotechnik und Elektronik 2: Elektromagnetische Felder und ihre Anwendungen },
     month         = { January },
     year          = { 2019 },
     edition       = { 2. Auflage },
     publisher     = { Springer },
     isbn          = { 978-3662582206 }
}

@misc { PTB19,
     author        = { Physikalisch-Technische Bundesanstalt PTB },
     title         = { <a href="https://www.ptb.de/cms/fileadmin/internet/forschung_entwicklung/countdown_new_si/Lesezeichen_zum_neuen_SI.pdf">Neue Definitionen im Internationalen Einheitensystem (SI)</a> },
     month         = { January },
     year          = { 2019 },
     note          = { PDF Download (accessed: 06.10.2022) }
}

@mastersthesis { Spre2018,
     author        = { Mawe Sprenger },
     title         = { Convolutional Neural Network for Depth and Odometry Estimation from Monocular Video },
     month         = { May },
     year          = { 2019 },
     school        = { Paderborn University },
     type          = { Master's thesis },
     abstract      = { Three-dimensional maps are an important source of information for the safe navigation of self-driving cars or autonomous robots. To create such maps, two data sources are required: On the one hand, continuous depth measurements have to be taken, while on the other hand the current position of the sensor has to be recorded to fuse all local measurements into a global map. However, exact measurements of depth and positional data are only possible with expensive equipment. To overcome this problem, the goal of this thesis is the development and implementation of a system for the simultaneous estimation of depth and camera ego-motion from monocular video streams. For this purpose, a convolutional neural network will be trained using a novel unsupervised learning method that enables learning of absolute scale depth and odometry without using ground-truth pose labels. To get depth and pose estimates that are as accurate as possible, different aspects of the developed learning method and the implemented system will be tested and evaluated for their relevancy. Since the quality of pose estimation depends significantly on an accurate depth estimation, methods which make use of temporal depth cues are tested for possible improvements. The developed system is evaluated regarding its performance and the estimated depth and odometry data is used to create 3D maps. }
}

@mastersthesis { Sti2018,
     author        = { Georg Stilow },
     title         = { Auswahl und Erweiterung eines lernbasierten Kantendetektionsverfahrens zur echtzeitf{"a}higen Objekterkennung },
     month         = { January },
     year          = { 2019 },
     school        = { Paderborn University },
     type          = { Master's thesis },
     abstract      = { Im Rahmen einer Objekterkennung werden Einzelbilder oder Bildsequenzen systematisch nach einem oder mehreren Objekten von Interesse durchsucht. Ein verbreiteter Ansatz zur Objekterkennung basiert auf der Analyse von Objektkonturen, die beispielsweise aus Kantenbildern extrahiert werden. Kantendetektionsverfahren liefern jedoch zun{"a}chst mehrdeutige und oftmals unterbrochene Konturen in Form von Grauwertbildern. In dieser Arbeit werden lernbasierte Kantendetektionsverfahren untersucht und anhand einer Gegen{"u}berstellung ein Verfahren zur Erkennung von Objektkonturen ausgew{"a}hlt. Das ausgew{"a}hlte Verfahren muss zun{"a}chst trainiert werden, wozu zum Beispiel ein Datensatz verwendet werden kann, der aus nat{"u}rlichen Bildern und den von Testpersonen erstellten Segmentierungen besteht. Nach dem Training werden Grauwertbilder erstellt, deren Helligkeitswerte die Wahrscheinlichkeit f{"u}r eine dort positionierte Kante repr{"a}sentieren. Um die Grauwertbilder in Objekterkennungsverfahren verwenden zu k{"o}nnen, sollen Mehrdeutigkeiten aufgel{"o}st und unterbrochene Konturen verbunden werden, so dass sich geeignete bin{"a}re Kantenbilder ergeben. Das resultierende Gesamtverfahren soll dabei m{"o}glichst echtzeitf{"a}hig sein. }
}

@article { BORN2019,
     author        = { Jan T{"u}nnermann and Christian Born and B{"a}rbel Mertsching },
     title         = { Saliency from Growing Neural Gas: Learning Pre-attentional Structures for a Flexible Attention System },
     month         = { September },
     year          = { 2019 },
     journal       = { IEEE Transactions on Image Processing },
     number        = { 11 },
     pages         = { 5296 -- 5307 },
     volume        = { 28 },
     issn          = { 1941-0042  },
     abstract      = { Artificial visual attention has been an active research area for over two decades. Especially, the concept of saliency has been implemented in many different ways. Early approaches aimed at closely modeling saliency processing with concepts from biological attention to provide (at least in the long run) general-purpose attention for technical systems. More recent approaches have departed from this agenda, turning to more specific attention-guided tasks, most notably the accurate extraction of salient objects, for which large-scale ground truth datasets make it possible to quantify progress. While the first type of models is troubled by weak performance in these specific tasks, the second type, as we show with a new benchmark, has lost the ability to predict saliency in the original sense, which may be an important factor for future general-purpose attention systems. Here, we describe a new approach using growing neural gas to obtain pre-attentional structures for a scene at an early processing stage. On this basis, traditional saliency concepts can be applied while at the same time they can be linked to mechanisms that make models successful in salient object detection. The model shows high performance at predicting traditional saliency and makes substantial progress toward salient object detection, although it cannot reach the top-level performance of some specialized methods. We discuss the important implications of our findings. }
}