Refine
Document Type
- Doctoral Thesis (18) (remove)
Keywords
Institute
- Institut für Computervisualistik (18) (remove)
Human action recognition from a video has received growing attention in computer vision and has made significant progress in recent years. Action recognition is described as a requirement to decide which human actions appear in videos. The difficulties involved in distinguishing human actions are due to the high complexity of human behaviors as well as appearance variation, motion pattern variation, occlusions, etc. Many applications use human action recognition on captured video from cameras, resulting in video surveillance systems, health monitoring, human-computer interaction, and robotics. Action recognition based on RGB-D data has increasingly drawn more attention to it in recent years. RGB-D data contain color (Red, Green, and Blue (RGB)) and depth data that represent the distance from the sensor to every pixel in the object (object point). The main problem that this thesis deals with is how to automate the classification of specific human activities/actions through RGB-D data. The classification process of these activities utilizes a spatial and temporal structure of actions. Therefore, the goal of this work is to develop algorithms that can distinguish these activities by recognizing low-level and high-level activities of interest from one another. These algorithms are developed by introducing new features and methods using RGB-D data to enhance the detection and recognition of human activities. In this thesis, the most popular state-of-the-art techniques are reviewed, presented, and evaluated. From the literature review, these techniques are categorized into hand-crafted features and deep learning-based approaches. The proposed new action recognition framework is based on these two categories that are approved in this work by embedding novel methods for human action recognition. These methods are based on features extracted from RGB-D data that are
evaluated using machine learning techniques. The presented work of this thesis improves human action recognition in two distinct parts. The first part focuses on improving current successful hand-crafted approaches. It contributes into two significant areas of state-of-the-art: Execute the existing feature detectors, and classify the human action in the 3D spatio-temporal domains by testing a new combination of different feature representations. The contributions of this part are tested based on machine learning techniques that include unsupervised and supervised learning to evaluate this suitability for the task of human action recognition. A k-means clustering represents the unsupervised learning technique, while the supervised learning technique is represented by: Support Vector Machine, Random Forest, K-Nearest Neighbor, Naive Bayes, and Artificial Neural Networks classifiers. The second part focuses on studying the current deep-learning-based approach and how to use it with RGB-D data for the human action recognition task. As the first step of each contribution, an input video is analyzed as a sequence of frames. Then, pre-processing steps are applied to the video frames, like filtering and smoothing methods to remove the noisy data from each frame. Afterward, different motion detection and feature representation methods are used to extract features presented in each frame. The extracted features
are represented by local features, global features, and feature combination besides deep learning methods, e.g., Convolutional Neural Networks. The feature combination achieves an excellent accuracy performance that outperforms other methods on the same RGB-D datasets. All the results from the proposed methods in this thesis are evaluated based on publicly available datasets, which illustrate that using spatiotemporal features can improve the recognition accuracy. The competitive experimental results are achieved overall. In particular, the proposed methods can be better applied to the test set compared to the state-of-the-art methods using the RGB-D datasets.
Efficient Cochlear Implant (CI) surgery requires prior knowledge of the cochlea’s size and its characteristics. This information helps to select suitable implants for different patients. Registered and fused images helps doctors by providing more informative image that takes advantages of different modalities. The cochlea’s small size and complex structure, in addition to the different resolutions and head positions during imaging, reveals a big challenge for the automated registration of the different image modalities. To obtain an automatic measurement of the cochlea length and the volume size, a segmentation method of cochlea medical images is needed. The goal of this dissertation is to introduce new practical and automatic algorithms for the human cochlea multi-modal 3D image registration, fusion, segmentation and analysis. Two novel methods for automatic cochlea image registration (ACIR) and automatic cochlea analysis (ACA) are introduced. The proposed methods crop the input images to the cochlea part and then align the cropped images to obtain the optimal transformation. After that, this transformation is used to align the original images. ACIR and ACA use Mattes mutual information as similarity metric, the adaptive stochastic gradient descent (ASGD) or the stochastic limited memory Broyden–Fletcher–Goldfarb–Shanno (s-LBFGS) optimizer to estimate the parameters of 3D rigid transform. The second stage of nonrigid registration estimates B-spline coefficients that are used in an atlas-model-based segmentation to extract cochlea scalae and the relative measurements of the input image. The image which has segmentation is aligned to the input image to obtain the non-rigid transformation. After that the segmentation of the first image, in addition to point-models are transformed to the input image. The detailed transformed segmentation provides the scala volume size. Using the transformed point-models, the A-value, the central scala lengths, the lateral and the organ of corti scala tympani lengths are computed. The methods have been tested using clinical 3D images of total 67 patients: from Germany (41 patients) and Egypt (26 patients). The atients are of different ages and gender. The number of images used in the experiments is 217, which are multi-modal 3D clinical images from CT, CBCT, and MRI scanners. The proposed methods are compared to the state of the arts ptimizers related medical image registration methods e.g. fast adaptive stochastic gradient descent (FASGD) and efficient preconditioned tochastic gradient descent (EPSGD). The comparison used the root mean squared distance (RMSE) between the ground truth landmarks and the resulted landmarks. The landmarks are located manually by two experts to represent the round window and the top of the cochlea. After obtaining the transformation using ACIR, the landmarks of the moving image are transformed using the resulted transformation and RMSE of the transformed landmarks, and at the same time the fixed image landmarks are computed. I also used the active length of the cochlea implant electrodes to compute the error aroused by the image artifact, and I found out an error ranged from 0.5 mm to 1.12 mm. ACIR method’s RMSE average was 0.36 mm with a standard deviation (SD) of 0.17 mm. The total time average required for registration of an image pair using ACIR was 4.62 seconds with SD of 1.19 seconds. All experiments are repeated 3 times for justifications. Comparing the RMSE of ACIR2017 and ACIR2020 using paired T-test shows no significant difference (p-value = 0.17). The total RMSE average of ACA method was 0.61 mm with a SD of 0.22 mm. The total time average required for analysing an image was 5.21 seconds with SD of 0.93 seconds. The statistical tests show that there is no difference between the results from automatic A-value method and the manual A-value method (p-value = 0.42). There is no difference also between length’s measurements of the left and the right ear sides (p-value > 0.16). Comparing the results from German and Egypt dataset shows there is no difference when using manual or automatic A-value methods (p-value > 0.20). However, there is a significant difference when using ACA2000 method between the German and the Egyptian results (p-value < 0.001). The average time to obtain the segmentation and all measurements was 5.21 second per image. The cochlea scala tympani volume size ranged from 38.98 mm3 to 57.67 mm3 . The combined scala media and scala vestibuli volume size ranged from 34.98 mm 3 to 49.3 mm 3 . The overall volume size of the cochlea should range from 73.96 mm 3 to 106.97 mm 3 . The lateral wall length of scala tympani ranged from 42.93 mm to 47.19 mm. The organ-of-Corti length of scala tympani ranged from 31.11 mm to 34.08 mm. Using the A-value method, the lateral length of scala tympani ranged from 36.69 mm to 45.91 mm. The organ-of-Corti length of scala tympani ranged from 29.12 mm to 39.05 mm. The length from ACA2020 method can be visualised and has a well-defined endpoints. The ACA2020 method works on different modalities and different images despite the noise level or the resolution. In the other hand, the A-value method works neither on MRI nor noisy images. Hence, ACA2020 method may provide more reliable and accurate measurement than the A-value method. The source-code and the datasets are made publicly available to help reproduction and validation of my result.
Das Hauptziel der vorliegenden Arbeit ist die Absicherung der Qualität eines pharmazeutischen Produktionsprozesses durch die Überprüfung des Volumens mikroskopischer Polymerstäbchen mit einem hochgenauen 3D Messverfahren. Die Polymerstäbchen werden für pharmazeutische Anwendungen hergestellt. Aus Gründen der Qualitätssicherung muss das Istgewicht überprüft werden. Derzeit werden die Polymerstäbchen stichprobenartig mit einer hochpräzisen Waage gewogen. Für die nächste Generation von Polymeren wird angenommen, dass die Produktabmessungen weiter reduziert werden sollen und die Produktionstoleranzen auf 2,5% gesenkt werden. Die daraus resultierenden Genauigkeitsanforderungen übersteigen jedoch die Möglichkeiten der Wiegetechnik. Bei homogenen Materialien ist die Masse proportional zum Volumen. Aus diesem Grund kommt dessen Bestimmung als Alternative in Frage. Dies verschafft Zugang zu optischen Messverfahren und deren Flexibilität und Genauigkeitpotenzial. Für den Entwurf eines auf die Fragestellung angepassten Messkonzeptes sind weiterhin von Bedeutung, dass das Objekt kontaktlos, mit einer Taktzeit von maximal fünf Sekunden vermessen und das Volumen approximiert wird. Die Querschnitte der Polymerstäbchen sind etwa kreisförmig. Aufgrund der Herstellung der Fragmente kann nicht davon ausgegangen werden, dass die Anlageflächen orthogonal zur Symmetrieachse des Objektes sind. Daher muss analysiert werden, wie sich kleine Abweichungen von kreisförmigen Querschnitten sowie die nicht idealen Anlageflächen auswirken. Die maximale Standardabweichung für das Volumen, die nicht überschritten werden sollte, beträgt 2,5%. Dies entspricht einer maximalen Abweichung der Querschnittsfläche um 1106 µm² (Fehlerfortpfanzung). Als Bewertungskriterium wird der Korrelationskoeffzient zwischen den gemessenen Volumina und den Massen bestimmt. Ein ideales Ergebnis wäre 100%. Die Messung zielt auf einen Koeffzienten von 98% ab. Um dies zu erreichen, ist ein präzises Messverfahren für Volumen erforderlich. Basierend auf dem aktuellen Stand der Technik können die vorhandenen optischen Messverfahren nicht verwendet werden. Das Polymerstäbchen wird von einer Kamera im Durchlicht beobachtet. Daher sind der Durchmesser und die Länge sichtbar. Das Objekt wird mittels einer mechanischen Vorrichtung um die Längsachse gedreht. So können Bilder von allen Seiten aufgenommen werden. Der Durchmesser und die Länge werden mit der Bildverarbeitung berechnet. Das neue Konzept vereint die Vorteile der Verfahren: Es ist unempfindlich gegen Farb-/Helligkeitsänderungen und die Bilder können in beliebiger Anzahl aufgenommen werden. Außerdem sind die Erfassung und Auswertung wesentlich schneller. Es wird ein Entwurf und die Umsetzung einer Lösung zur hochpräzisen Volumenmessung von Polymerstäbchen mit optischer Messtechnik und Bildverarbeitung ausgearbeitet. Diese spezielle Prozesslösung in der Prozesslinie (inline) sollte eine 100%ige Qualitätskontrolle während der Produktion garantieren. Die Zykluszeiten des Systems sollte fünf Sekunden pro Polymerstäbchen nicht überschreiten. Die Rahmenbedienungen für den Prozess sind durch die Materialeigenschaften des Objekts, die geringe Objektgröße (Breite = 199 µm, Länge = 935 µm bis 1683 µm) und die undeffinierte Querschnittsform (durch den Trocknungsprozess) vorgegeben. Darüber hinaus sollten die Kosten für den Prozess nicht zu hoch sein. Der Messaufbau sollte klein sein und ohne Sicherheitsvorkehrungen oder Abschirmungen arbeiten. Das entstandene System nimmt die Objekte in verschiedenen Winkelschritten auf, wertet mit Hilfe der Bildverarbeitung die Aufnahmen aus und approximiert das Volumen. Der Korrelationskoffizient zwischen Volumen und Gewicht beträgt für 77 Polymerstäbchen mit einem Gewicht von 37 µg bis 80 µg 99; 87%. Mit Hilfe eines Referenzsystems kann die Genauigkeit der Messung bestimmt werden. Die Standardabweichung sollte maximal 2,5% betragen. Das entstandene System erzielt eine maximale Volumenabweichung von 1,7%. Die Volumenvermessung erfüllt alle Anforderungen und kann somit als Alternative für die Waage verwendet werden.
This thesis presents the analysis of gamebased touristic applications. In tourism, actions can only be motivated intrinsic. Thus, this thesis at first researches specific intrinsic motivation concepts. It shows how gamebased motivation can be produced on purpose and answers the question whether gamebased motivation can be transferred to non-gamebased applications.
Using these results, different touristic applications have been developed and evaluated.
All applications aimed to add value to the touristic experience. The applications are sorted by their mobility. There are completely mobile, completely stationary and hybrid systems in this work. There are different ways to add value which are presented in this work: Gamebased exploration, knowledge transfer and social interaction between tourists.
Finally, an authoring tool for gamebased touristic tours on smartphones is presented.
Augmented reality (AR) applications typically extend the user's view of the real world with virtual objects.
In recent years, AR has gained increasing popularity and attention, which has led to improvements in the required technologies. AR has become available to almost everyone.
Researchers have made great progress towards the goal of believable AR, in which the real and virtual worlds are combined seamlessly.
They mainly focus on issues like tracking, display technologies and user interaction, and give little attention to visual and physical coherence when real and virtual objects are combined. For example, virtual objects should not only respond to the user's input; they should also interact with real objects. Generally, AR becomes more believable and realistic if virtual objects appear fixed or anchored in the real scene, appear indistinguishable from the real scene, and response to any changes within it.
This thesis examines on three challenges in the field of computer vision to meet the goal of a believable combined world in which virtual objects appear and behave like real objects.
Firstly, the thesis concentrates on the well-known tracking and registration problem. The tracking and registration challenge is discussed and an approach is presented to estimate the position and viewpoint of the user so that virtual objects appear fixed in the real world. Appearance-based line models, which keep only relevant edges for tracking purposes, enable absolute registration in the real world and provide robust tracking. On the one hand, there is no need to spend much time creating suitable models manually. On the other hand, the tracking can deal with changes within the object or the scene to be tracked. Experiments have shown that the use of appearance-based line models improves the robustness, accuracy and re-initialization speed of the tracking process.
Secondly, the thesis deals with the subject of reconstructing the surface of a real environment and presents an algorithm to optimize an ongoing surface reconstruction. A complete 3D surface reconstruction of the target scene
offers new possibilities for creating more realistic AR applications. Several interactions between real and virtual objects, such as collision and occlusions, can be handled with physical correctness. Whereas previous methods focused on improving surface reconstructions offline after a capturing step, the presented method de-noises, extends and fills holes during the capturing process. Thus, users can explore an unknown environment without any preparation tasks such as moving around and scanning the scene, and without having to deal with the underlying technology in advance. In experiments, the approach provided realistic results where known surfaces were extended and filled in plausibly for different surface types.
Finally, the thesis focuses on handling occlusions between the real and virtual worlds more realistically, by re-interpreting the occlusion challenge as an alpha matting problem. The presented method overcomes limitations in state-of-the-art methods by estimating a blending coefficient per pixel of the rendered virtual scene, instead of calculating only their visibility. In several experiments and comparisons with other methods, occlusion handling through alpha matting worked robustly and overcame limitations of low-cost sensor data; it also outperformed previous work in terms of quality, realism and practical applicability.
The method can deal with noisy depth data and yields realistic results in regions where foreground and background are not strictly separable (e.g. caused by fuzzy objects or motion blur).
Studies in recent years have demonstrate adolescents and young adults to have a deficient data protection competence, however children and adolescents between the ages of ten and 13 were mostly not focus of these studies. Therefore, the guiding question of the work is how data protection competence is developed in children and adolescents at a young age in order to be able to infer suitable, educational concepts for this age group. At the beginning of the work, a data protection competence model is derived from a media competence model, which serves as the basis for the further field investigation. A survey was carried out at general secondary schools in Rhineland-Palatinate, which shows that the respondents still have sufficiently developed Risk Assessment Competence, but were insufficiently developed in terms of knowledge, Selection and Usage Competence and the Implementation Competence. Recommendations for actions are given in the last part of the work – containing learning goal descriptions to be possibly implemented in an educational framework – in order to address this issue.
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.
Technologische Fortschritte auf dem Gebiet der integrierten Halbleitertechnik, die unter anderem auch zur gestiegenen Leistungsfähigkeit der Kamerasensoren beitragen, konzentrierten sich bisher primär auf die Schnelligkeit und das Auflösungsvermögen der Sensoren. Die sich ständig verändernde Entwicklung hat jedoch direkte Folgen auf das physikalische Verhalten einer Kamera und damit auch Konsequenzen für die erreichbare geometrische Genauigkeit einer photogrammetrischen 3D-Rekonstruktion. Letztere stand bisher nicht im Fokus der Forschung und ist eine Aufgabe, der sich diese Arbeit im Sinne der Photogrammetrie und Messtechnik stellt. Aktuelle Untersuchungen und Erfahrungen aus industriellen Projekten zeigen in diesem Zusammenhang, dass das geometrisch-physikalische Verhalten digitaler Kameras - für höchste photogrammetrische Ansprüche - noch nicht ausreichend modelliert ist. Direkte Aussagen zur erreichbaren Genauigkeit bei gegebener Hardware erweisen sich daher bislang als unzureichend. Ferner kommt es aufgrund der unpräzisen Modellierung zu Einbußen in der Zuverlässigkeit der erreichten Ergebnisse. Für den Entwickler präziser kamerabasierter Messverfahren folgt daraus, dass zu einer optimalen Schätzung der geometrischen Genauigkeit und damit auch vollständigen Ausschöpfung der Messkamera geeignete mathematische Modelle erforderlich sind, die das geometrisch physikalische Verhalten bestmöglich beschreiben. Diese Arbeit beschreibt, wie die erreichbare Genauigkeit einer Bündelblockausgleichung, schon a priori mithilfe des EMVA1288 Standards approximiert werden kann. Eine in diesem Zusammenhang wichtige Teilaufgabe ist die Schaffung einer optimalen Messanordnung. Hierzu gehören Untersuchungen der üblicherweise verwendeten Kalibrierkörper und die Beseitigung von systematischen Fehlern vor und nach der Bündelblockausgleichung. Zum Nachweis dieser Systematiken wird eine auf statistischem Lernen basierende Methode beschrieben und untersucht. Erst wenn alle genauigkeitsmindernden Einflüsse berücksichtigt sind, wird der Anteil des Sensors in den Messdaten sichtbar und damit auch mathematisch parametrisierbar. Die Beschreibung des Sensoreinflusses auf die erreichbare Genauigkeit der Bündelblockausgleichung erfolgt in drei Schritten. Der erste Schritt beschreibt den Zusammenhang zwischen ausgewählten EMVA1288-Kennzahlen und der Unsicherheit eines Grauwertes. Der zweite Schritt ist eine Modellierung dieser Grauwertunsicherheit als Zentrumsunsicherheit einer Zielmarke. Zur Beschreibung dieser Unsicherheit innerhalb der Bündelblockausgleichung wird ein stochastisches Modell, basierend auf dem EMVA1288-Standard, vorgeschlagen. Ausgehend vom Rauschen des Zielmarkenmittelpunktes wird im dritten Schritt die Unsicherheit im Objektraum beispielhaft mit Hilfe von physikalisch orientierten Simulationen approximiert. Die Wirkung der vorgeschlagenen Methoden wird anhand von Realkalibrierungen nachgewiesen. Abschließend erfolgt die Diskussion der vorgeschlagenen Methoden und erreichten Ergebnisse sowie ein Ausblick auf kommende Untersuchungen.
The cytological examination of bone marrow serves as clarification of variations in blood smears. It is also used for the clarification of anemia, as exclusion of bone marrow affection at lymphoma and at suspicion of leukemia. The morphological evaluation of hematopoietic cells is the basis for the creation of the diagnosis and for decision support for further diagnostics. Even for experienced hematologists the manual classification of hematopoietic cells is time-consuming, error-prone and subjective. For this reason new methods in the field of image processing and pattern recognition for the automatic classification including preprocessing steps are developed for a computer-assisted microscopy system. These methods are evaluated by means of a huge reference database. The proposed image analysis procedures comprise methods for the automated detection of smears, for the determination of relevant regions, for the localization and segmentation of single hematopoietic cells as well as for the feature extraction and classification task. These methods provide the basis for the first system for the automated, morphological analysis of bone marrow aspirates for leukemia diagnosis and are therefore a major contribution for a better and more efficient patient care in the future.
The goal of this thesis is the development of methods for augmented image synthesis using 3D photo collections. 3D photo collections are representations of real scenes automatically generated from single photos and describe a scene as a set of images with known camera poses as well as a sparse point-based model of the scene geometry. The main goal is to perform a photo-realistic augmented image synthesis of real and virtual parts, where the real scene is provided as a 3D photo collection. Therefore, three main problems are addressed.
Since the photos may be represented in different device-specific RGB color spaces, a color characterization of the 3D photo collections is necessary to gain correct color information that is consistent with human perception. The proposed novel method automatically transforms all images into a common RGB color space and thereby simplifies color characterization of 3D photo collections.
As a main problem for augmented image synthesis, all environmental lighting has to be known in order to apply illumination to virtual parts that is consistent with the real portions shown in the photos. To solve this problem, two novel methods were developed to reconstruct the lighting from 3D photo collections.
In order to perform image synthesis for arbitrary views on the scene, an image-based approach was developed that generates new views in 3D photo collections making direct use of its point cloud. The novel method creates new views in real-time and allows free-navigation.
In conclusion, the proposed novel methods show that 3D photo collections are a useful representation for real scenes in Augmented Reality and they can be used to perform a realistic image synthesis of real and virtual portions.