Refine
Document Type
- Doctoral Thesis (18) (remove)
Keywords
Institute
- Institut für Computervisualistik (18) (remove)
In the context of augmented reality we define tracking as a collection of methods to obtain the position and orientation (pose) of a user. By means of various displaying techniques, this ensures a correct visual overlay of graphical information onto the reality perceived. Precise results for calculation of the camera pose are gained by methods of image processing, usually analyzing the pixels of an image and extracing features, which can be recognized over the image sequence. However, these methods do not regard the process of image synthesis or at least in a very simplyfied way. In contrast, the class of model-based methods assumes a given 3D model of the observed scene. Based on the model data features can be identified to establish correspondences in the camera image. From these feature correspondences the camera pose is calculated. An interesting approach is the strategy of analysis-by-synthesis, regarding the computer graphics rendering process for extending the knowledge about the model by information from image synthesis and other environment variables.
In this thesis the components of a tracking system are identified and further it is analyzed, to what extend information about the model, the rendering process and the environment can contribute to the components for improvement of the tracking process using analysis-by-synthesis. In particular, by using knowledge as topological information, lighting or perspective, the feature synthesis and correspondence finding should lead to visually unambiguous features that can be predicted and evaluated to be suitable for stable tracking of the camera pose.
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).
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.
In this work has been examined, how the existing model of the simulation of cables and hoses can be advanced. Therefore an investigation has been made on the main influences to the shape simulation and the factors of constraints and side conditions were analyzed. For the validation of the accuracy, the simulation has to be compared to real specimen behavior. To obtain a very precise digitalization of the shape, the choice was made to use a laser scanner that converts the pointcloud into a .vrml file which can be imported into the simulation environment. The assumption was that the simulation method itself has the highest impact to the simulated shape. This is why the capabilities of the most sophisticated methods have been analyzed. The main criterion for the success of a simulation approach proved not to be accuracy, as expected. Process integration and usability showed to be of higher interest for the efficient exertion. Other factors like the pricing, the functionality and the real-time capability were assayed as well. The analyzed methods are based on the solution of the equations of elasticity with different ways of discetization, finite-elements and a spring-impulse-system. Since the finite-element-system takes several minutes for the computation of the shape and the spring-impulse-system reacts retarded on user manipulation, the competitiveness of these approaches is low. The other methods distinguish more in real-time performance, data interfaces and functionality than in accuracy. For the accuracy of a system, the consideration of other factors proved to be very important. As one of these main factors, the accurate assignment of the material properties was indicated. Until the start of this work, only the finite-element-approach dealt with this factor, but no documentation or validation is provided. In the publications of the other methods, the material properties are estimated to obtain a plausible simulation shape. Therefore the specific material values of real specimen have been measured and assigned to the simulation. With the comparison to the real shape it has been proven that the accuracy is very high with the measured properties. Since these measurements are very costly and time consuming, an investigation on a faster and cheaper way to obtain these values has been made. It has been assumed that with the knowledge of the cross-section it should be possible to compute the specimen behavior. Since the braid distribution changes individually from specimen to specimen, a more general way to obtain the values needed to be found. The program composer has been developed, where only the number of the different braids and the taping is entered. It computes with very high precision the stiffness, the density and the final diameter of the bundle. With the measured values and the fitting to the real shape it has been proven that the simulation approach reflects the precise behavior of cables and hoses. Since the stiffness of the single braids is wasteful to measure, a measurement setup was created where the stiffness has a large impact to the shape. With known density, the stiffness of the specimen can be reconstructed precisely. Thus a fast and beneficial way of obtaining the stiffness of a cable has been invented. The poissons ratio of cables and bundles cannot be measured with a tensile test, since the inner structure is very complex. For hoses, the variation of the inner diameter has been measured during the tensile test as well. The resulting values were reasonable, but their accuracy could not be proven. For cables and hoses, it has been tried to obtain the poissons ratio via the computation of the cross section, but the influence of individual changes and the crosstalk of the braids is very high. Therefore a setup was constructed where the torsion stiffness can be measured. For cables and hoses, the individual cross-sections and taping lead to varying results. For hoses, expected and repeatable good values for the poissons ratio were obtained. The low influence of the poisons ratio in the range between 0 and 0.5 has been proven. Therefore we decided to follow the advice of [Old06] and our own experiences to set the poisons ratio for cables and bundles to 0.25. With the knowledge of the measurability and the capabilities of the developed program composer, a procedure to obtain material properties for bundles has been designed. 1. Measurement of the braid density with via pyknometer or mass, length and diameter. 2. Empirical reconstruction of the stiffness with the designed setup. 3. Composing the bundle with the program composer. 4. Adding a factor for the taping and transfer the values to the simulation. The model of the cable simulation has been improved as follows: The main influences in the simulation of cables and hoses are the simulation method, the material properties and the geometric constraints. To obtain higher accuracy, an investigation on the correct material properties is indispensable. The scientific determination of material properties for the simulation of cables, bundles and hoses has been performed for the first time. The influence of geometrical constraints has been analyzed and documented. The next steps are the analysis of pre-deformation and further investigations to the determination of the poisons ratio with a more precise torsion test. All analysis were made with the simulation approach fleXengine. A comparison to other simulation methods would be of high interest.
On the recognition of human activities and the evaluation of its imitation by robotic systems
(2023)
This thesis addresses the problem of action recognition through the analysis of human motion and the benchmarking of its imitation by robotic systems.
For our action recognition related approaches, we focus on presenting approaches that generalize well across different sensor modalities. We transform multivariate signal streams from various sensors to a common image representation. The action recognition problem on sequential multivariate signal streams can then be reduced to an image classification task for which we utilize recent advances in machine learning. We demonstrate the broad applicability of our approaches formulated as a supervised classification task for action recognition, a semi-supervised classification task for one-shot action recognition, modality fusion and temporal action segmentation.
For action classification, we use an EfficientNet Convolutional Neural Network (CNN) model to classify the image representations of various data modalities. Further, we present approaches for filtering and the fusion of various modalities on a representation level. We extend the approach to be applicable for semi-supervised classification and train a metric-learning model that encodes action similarity. During training, the encoder optimizes the distances in embedding space for self-, positive- and negative-pair similarities. The resulting encoder allows estimating action similarity by calculating distances in embedding space. At training time, no action classes from the test set are used.
Graph Convolutional Network (GCN) generalized the concept of CNNs to non-Euclidean data structures and showed great success for action recognition directly operating on spatio-temporal sequences like skeleton sequences. GCNs have recently shown state-of-the-art performance for skeleton-based action recognition but are currently widely neglected as the foundation for the fusion of various sensor modalities. We propose incorporating additional modalities, like inertial measurements or RGB features, into a skeleton-graph, by proposing fusion on two different dimensionality levels. On a channel dimension, modalities are fused by introducing additional node attributes. On a spatial dimension, additional nodes are incorporated into the skeleton-graph.
Transformer models showed excellent performance in the analysis of sequential data. We formulate the temporal action segmentation task as an object detection task and use a detection transformer model on our proposed motion image representations. Experiments for our action recognition related approaches are executed on large-scale publicly available datasets. Our approaches for action recognition for various modalities, action recognition by fusion of various modalities, and one-shot action recognition demonstrate state-of-the-art results on some datasets.
Finally, we present a hybrid imitation learning benchmark. The benchmark consists of a dataset, metrics, and a simulator integration. The dataset contains RGB-D image sequences of humans performing movements and executing manipulation tasks, as well as the corresponding ground truth. The RGB-D camera is calibrated against a motion-capturing system, and the resulting sequences serve as input for imitation learning approaches. The resulting policy is then executed in the simulated environment on different robots. We propose two metrics to assess the quality of the imitation. The trajectory metric gives insights into how close the execution was to the demonstration. The effect metric describes how close the final state was reached according to the demonstration. The Simitate benchmark can improve the comparability of imitation learning approaches.
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.
This thesis addresses the automated identification and localization of a time-varying number of objects in a stream of sensor data. The problem is challenging due to its combinatorial nature: If the number of objects is unknown, the number of possible object trajectories grows exponentially with the number of observations. Random finite sets are a relatively new theory that has been developed to derive at principled and efficient approximations. It is based around set-valued random variables that contain an unknown number of elements which appear in arbitrary order and are themselves random. While extensively studied in theory, random finite sets have not yet become a leading paradigm in practical computer vision and robotics applications. This thesis explores random finite sets in visual tracking applications. The first method developed in this thesis combines set-valued recursive filtering with global optimization. The problem is approached in a min-cost flow network formulation, which has become a standard inference framework for multiple object tracking due to its efficiency and optimality. A main limitation of this formulation is a restriction to unary and pairwise cost terms. This circumstance makes integration of higher-order motion models challenging. The method developed in this thesis approaches this limitation by application of a Probability Hypothesis Density filter. The Probability Hypothesis Density filter was the first practically implemented state estimator based on random finite sets. It circumvents the combinatorial nature of data association itself by propagation of an object density measure that can be computed efficiently, without maintaining explicit trajectory hypotheses. In this work, the filter recursion is used to augment measurements with an additional hidden kinematic state to be used for construction of more informed flow network cost terms, e.g., based on linear motion models. The method is evaluated on public benchmarks where a considerate improvement is achieved compared to network flow formulations that are based on static features alone, such as distance between detections and appearance similarity. A second part of this thesis focuses on the related task of detecting and tracking a single robot operator in crowded environments. Different from the conventional multiple object tracking scenario, the tracked individual can leave the scene and later reappear after a longer period of absence. Therefore, a re-identification component is required that picks up the track on reentrance. Based on random finite sets, the Bernoulli filter is an optimal Bayes filter that provides a natural representation for this type of problem. In this work, it is shown how the Bernoulli filter can be combined with a Probability Hypothesis Density filter to track operator and non-operators simultaneously. The method is evaluated on a publicly available multiple object tracking dataset as well as on custom sequences that are specific to the targeted application. Experiments show reliable tracking in crowded scenes and robust re-identification after long term occlusion. Finally, a third part of this thesis focuses on appearance modeling as an essential aspect of any method that is applied to visual object tracking scenarios. Therefore, a feature representation that is robust to pose variations and changing lighting conditions is learned offline, before the actual tracking application. This thesis proposes a joint classification and metric learning objective where a deep convolutional neural network is trained to identify the individuals in the training set. At test time, the final classification layer can be stripped from the network and appearance similarity can be queried using cosine distance in representation space. This framework represents an alternative to direct metric learning objectives that have required sophisticated pair or triplet sampling strategies in the past. The method is evaluated on two large scale person re-identification datasets where competitive results are achieved overall. In particular, the proposed method better generalizes to the test set compared to a network trained with the well-established triplet loss.
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
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.
This thesis focuses on the utilization of modern graphics hardware (GPU) for visualization and computation purposes, especially of volumetric data from medical imaging. The considerable increase in raw computing power in recent years has turned commodity systems into high-performance workstations. In combination with the direct rendering capabilities of graphics hardware, "visual computing" and "computational steering" approaches on large data sets have become feasible. In this regard several example applications and concepts such as the "ray textures" have been developed and are discussed in detail. As the amount of data to be processed and visualized is steadily increasing, memory and bandwidth limitations require compact representations of the data. While the compression of image data has been investigated extensively in the past, the thesis addresses possibilities of performing computations directly on the compressed data. Therefore, different categories of algorithms are identified and represented in the wavelet domain. By using special variants of the compressed format, efficient implementations of essential image processing algorithms are possible and demonstrate the potential of the approach. From the technical perspective, the GPU-based framework "Cascada" has been developed in the course of this thesis. The introduction of object-oriented concepts to shader programming, as well as a hierarchical representation of computation and/or visualization procedures led to a simplified utilization of graphics hardware while maintaining competitive performance. This is shown with different implementations throughout the contributions, as well as two clinical projects in the field of diagnosis assistance. On the one hand the semi-automatic segmentation of low-resolution MRI data sets of the human liver is evaluated. On the other hand different possibilities in assessing abdominal aortic aneurysms are discussed; both projects make use of graphics hardware. In addition, "Cascada" provides extensions towards recent general-purpose programming architectures and a modular design for future developments.