Patric Keller

Extracting and Visualizing Structural Features in Environmental Point Cloud LiDaR Data Sets

We present a user-assisted approach to extracting and visualizing structural features from point clouds obtained by terrestrial and airborne laser scanning devices. We apply a multi-scale approach to express the membership of local point environments to corresponding geometric shape classes in terms of probability. This information is filtered and combined to establish feature graphs which can be visualized in combination with the color-encoded feature and structural probability estimates of the measured raw point data. Our method can be used, for example, for exploring geological point data scanned from multiple viewpoints.

Modeling the effects of software on safety and reliability in complex embedded systems

The development of autonomous vehicle systems demands the increased usage of software based control mechanisms. Generally, this leads to very complex systems, whose proper functioning has to be ensured. In our work we aim at investigating and assessing the potential effects of software issues on the safety, reliability and availability of complex embedded autonomous systems. One of the key aspects of the research concerns the mapping of functional descriptions in form of integrated behavior-based control networks to State-Event Fault Tree models.

Reverse engineering with subdivision surfaces

Reverse engineering is concerned with the reconstruction of surfaces from three-dimensional point clouds originating from laser-scanned objects. We present an adaptive surface reconstruction method providing a hierarchy of quadrilateral meshes adapting surface topology when a mesh is refined. This way, a user can choose a model with proper resolution and topology from the hierarchy without having to run the algorithm multiple times with different parameters. The multiresolution mesh representation can be used subsequently for view-dependent rendering and wavelet compression.

THE EXTENDED STEREOSCOPIC HIGHLIGHTING TECHNIQUE FOR NODE-LINK DIAGRAMS: AN EMPIRICAL STUDY

The stereoscopic highlighting technique is a new emerging technique that supports using the depth cue in 3D devices as a highlighting technique for general nodelink representations. One of the main abilities of this technique is isolating the specific graph portions to magnify them for a detailed exploration without resorting other highlighting attributes such as color or shape. In this paper, we measure the accuracy of this technique by evaluating the ability of general users in reading the variation of depth values to encode some data aspects. To achieve this goal, we carried out a controlled user-study in which we asked users from different backgrounds to perform a specified set of tasks. The results show that using the depth cue in stereoscopic devices is readable for medium data sizes with at most three or four different layers. This can be useful in classifying the data set into a set of categories according to the depth value of the elemental layer.

Improving Safety-Critical Systems by Visual Analysis

The importance analysis provides a means of analyzing the contribution of potential low-level system failures to identify and assess vulnerabilities of safety-critical systems. Common approaches attempt to enhance the system safety by addressing vulnerabilities using an iterative analysis process, while considering relevant constraints, e.g., cost, for optimizing the improvements. Typically, data regarding the analysis process is presented across several views with few interactive associations among them. Consequently, this hampers the identification of meaningful information supporting the decision making process. In this paper, we propose a visualization system that visually supports engineers in identifying proper solutions. The visualization integrates a decision tree with a plot representing the cause-effect relationship between the improvement ideas of vulnerabilities and the resulting risk reduction of system. Associating a component fault tree view with the plot allows to maintain helpful context information. The introduced visualization approach enables system and safety engineers to identify and analyze optimal solutions facilitating the improvement of the overall system safety.

Visual approach facilitating the importance analysis of component fault trees

(Component) fault tree analysis is a safety analysis technique of embedded systems. Importance analysis estimates the respective contributions of potential basic failures to an overall system failure. The analysis results are typically represented in data-aggregated forms. There are only few associations between these forms and component fault tree structures that provide meaningful information. In this paper, we propose a visualization approach that integrates the importance analysis results with structures of component fault trees. This approach facilitates the identification of the critical components and supports the analysis of the influence of the important basic failures.

A General Introduction To Graph Visualization Techniques

Generally, a graph is an abstract data type used to represent relations among a given set of data entities. Graphs are used in numerous applications within the field of information visualization, such as VLSI (circuit schematics), state-transition diagrams, and social networks. The size and complexity of graphs easily reach dimensions at which the task of exploring and navigating gets crucial. Moreover, additional requirements have to be met in order to provide proper visualizations. In this context, many techniques already have been introduced. This survey aims to provide an introduction on graph visualization techniques helping the reader to gain a first insight into the most fundamental techniques. Furthermore, a brief introduction about navigation and interaction tools is provided.

Quality Improvement Through Visualization of Software and Systems

Many organizations still lack support for obtaining control over their system development processes and for determining the performance of their processes and the quality of the produced products. Systematic support for detecting and reacting to critical process and product states in order to achieve planned goals is often missing. As systems and software become bigger and more complex, classic approaches reach their limits, due to the difficulty of extracting relevant information from a large volume of measures. Here, suitable visualization and virtual reality solutions can offer a clear advantage by representing the relevant information in a more easily recognizable form. However, many resulting visualizations are still hard to understand, even for experts. This opens the door for researching modern, human-centered approaches that provide the user with visualization and interaction models for visually analyzing and understanding the underlying complex data. This chapter focuses on two main topics: system visualization and software visualization.

Surface Reconstruction from Unorganized 3D Point Clouds

Computer-based surfacemodels are indispensable in several fields of science and engineering. For example, the design and manufacturing of vehicles, such as cars and aircrafts, would not be possible without sophisticated CAD and simulation tools predicting the behavior of the product. On the other hand, designers often do not like working on virtual models, though sophisticated tools, like immersive VR-environments are available. Hence, a designer may produce a physical prototype made from materials of his choice that can be easily assembled and shaped like clay models. Reverse engineering is the process of reconstructing digital representations from physical models. The overall reverse-engineering framework mainly is composed of four steps (see Figure 1): data acquisition, pre-processing, surface reconstruction, and post-processing;

Construction of Implicit Surfaces from Point Clouds Using a Feature-based Approach

We present a novel feature-based approach to surface generation from point clouds in three-dimensional space obtained by terrestrial and airborne laser scanning. In a first step, we apply a multiscale clustering and classification of local point set neighborhoods by considering their geometric shape. Corresponding feature values quantify the similarity to curve-like, surface-like, and solid-like shapes. For selecting and extracting surface features, we build a hierarchical trivariate B-spline representation of this surface feature function. Surfaces are extracted with a variant of marching cubes (MC), providing an inner and outer shell that are merged into a single non-manifold surface component at the field’s ridges. By adapting the isovalue of the feature function the user may control surface topology and thus adapt the extracted features to the noise level of the underlying point cloud. User control and adaptive approximation make our method robust for noisy and complex point data.