Optical coherence tomography (OCT) enables noninvasive high-resolution imaging of the human retina and therefore, plays a fundamental role in detecting a wide range of ocular diseases. Yet, OCT data often vary in quality and show strong parameter dependencies. We propose a visual analysis approach to support users in understanding the influence of parameters on different aspects of the data. First, we outline the problem scope and derive requirements for a visual parameter analysis of OCT data. Second, we devise matched visual designs that disclose the impact of specific parameter values and the relationships between multiple parameter settings. With our systematic approach we aim at helping users in choosing suitable parameter settings and finding a balance between acquisition effort and data quality.

Beyond other domains, the field of immersive analytics makes use of Augmented Reality techniques to successfully support users in analyzing data. When displaying ubiquitous data integrated into the everyday life, spatial immersion issues like depth perception, data localization and object relations become relevant. Although there is a variety of techniques to deal with those, they are difficult to apply if the examined data or the reference space are large and abstract. In this work, we discuss observed problems in such immersive analytics systems and the applicability of current countermeasures to identify needs for action.

We present a novel depth image enhancement approach for RGB-D cameras such as the Kinect. Our approach employs optical flow of color images for refining the quality of corresponding depth images. We track every depth pixel over a sequence of frames in the temporal domain and use valid depth values of the same point for recovering missing and inaccurate information. We conduct experiments on different test datasets and present visually appealing results. Our method significantly reduces the temporal noise level and the flickering artifacts.

Exploring and comparing categorical time series and finding temporal patterns are complex tasks in the field of time series data mining. Although different analysis approaches exist, these tasks remain challenging, especially when numerous time series are considered at once. We propose a visual analysis approach that supports exploring such data by ordering time series in meaningful ways. We provide interaction techniques to steer the automated arrangement and to allow users to investigate patterns in detail.

Recognizing activities has become increasingly relevant in many application domains, such as security or ambient assisted living. To handle different scenarios, the underlying automated algorithms are configured using multiple input parameters. However, the influence and interplay of these parameters is often not clear, making exhaustive evaluations necessary. On this account, we propose a visual analytics approach to support users in understanding the complex relationships between parameters, recognized activities, and associated accuracies. First, representative parameter settings are determined. Then, the respective output is computed and statistically analyzed to assess parameters' influence in general. Finally, visualizing the parameter settings along with the activities provides overview and allows to investigate the computed results in detail. Coordinated interaction helps to explore dependencies, compare different settings, and examine individual activities. By integrating automated, visual, and interactive means users can select parameter values that meet desired quality criteria. We demonstrate the application of our solution in a use case with realistic complexity, involving a study of human protagonists in daily living with respect to hundreds of parameter settings.

Analyzing movements in their spatial and temporal context is a complex task. We are additionally interested in understanding the movements' dependency on parameters that govern the processes behind the movement. We propose a visual analytics approach combining analytic, visual, and interactive means to deal with the added complexity. The key idea is to perform an analytical extraction of features that capture distinct movement characteristics. Different parameter configurations and extracted features are then visualized in a compact fashion to facilitate an overview of the data. Interaction enables the user to access details about features, to compare features, and to relate features back to the original movement. We instantiate our approach with a repository of more than twenty accepted and novel features to help analysts in gaining insight into simulations of chaotic behavior of thousands of entities over thousands of data points. Domain experts applied our solution successfully to study dynamic groups in such movements in relation to thousands of parameter configurations.

Corneal confocal microscopy has enabled recent advances in the diagnosis of nerve damage for a range of peripheral neuropathies, in particular diabetic neuropathy. However, analyzing the multivariate microscopic values in association with other clinical and neurological data remains challenging. To this end, we propose a visual analysis approach to show the data in comprehensible manner and in this way, to support experts in understanding and assessing complex relationships.

Node-link and matrix representations are widely applied for graph exploration. However, when it comes to editing graphs, matrix representations are mostly neglected. In this work, we investigate the suitability of matrix representations especially for graph editing. Based on a review of the characteristics of matrices in terms of representation and interaction, we propose first techniques for edge-based editing tasks. We combine our matrix-based techniques with classic node-link solutions in an interactive prototype for touch enabled graph editing.

Reconstructing processes from measurements of multiple sensors over time is an important task in many application domains. For the reconstruction, these multivariate time-series can be automatically processed. However, the outcomes of automated algorithms often vary in quality and show strong parameter dependencies, making manual inspections and adjustments of the results necessary. We propose a visual analysis approach to support the user in understanding parameters' influences on these results. With our approach the user can identify and select parameter settings that meet certain quality criteria. The proposed visual and interactive design helps to identify relationships and temporal patterns, supports subsequent decision making, and promotes higher accuracy as well as confidence in the results.

The visualization of simulation trajectories is a well-established approach to analyze simulated processes. Likewise, the visualization of the parameter space that configures a simulation is a well-known method to get an overview of possible parameter combinations. This paper follows the premise that both of these approaches are actually two sides of the same coin; since the input parameters influence the simulation outcome, it is desirable to visualize and explore both in a combined manner. The main challenge posed by such an integrated visualization is the combinatorial explosion of possible parameter combinations. It leads to insurmountably high simulation runtimes and screen space requirements for their visualization. The Visual Analytics approach presented in this paper targets this issue by providing a visualization of a coarsely sampled subspace of the parameter space and its corresponding simulation outcome. In this visual representation, the analyst can identify regions for further drill-down and thus finer subsampling. We aid this identification by providing visual cues based on heterogeneity metrics. These indicate in which regions of the parameter space deviating behavior occurs at a more fine-grained scale and thus warrants further investigation and possible re-computation. We demonstrate our approach in the domain of systems biology by a visual analysis of a rule-based model of the canonical Wnt signaling pathway that plays a major role in embryonic development. In this case, the aim of the domain experts was to systematically explore the parameter space to determine those parameter configurations that match experimental data sufficiently well.

Diabetic neuropathy is the most common long-term complication of diabetes mellitus. It comes along with significant nerve dysfunction, which is not reversible. Hence, it is essential to detect nerve fibre abnormalities as early as possible. In this paper, we investigate markers describing degradation of corneal nerves. We apply statistical computations and visual analysis to identify those variables of two clinical studies that separate DN patients from a control group. In this way, the diagnosis of DN patients is supported. The visual analysis is based on different representations visualizing both the statistical results and the gathered multi-variate data. The user can interactively manipulate the views, or select data that will be shown by further displays. In this way, the understanding of the data and its classification is supported. Ambiguous categorisations can be identified and grouped into a so-called "fuzzy group". For this group, further investigations are needed to decide about diabetic neuropathy.

In this paper, we describe a step-wise approach to utilize ophthalmic markers for detecting early diabetic neuropathy (DN), the most common long-term complication of diabetes mellitus. Our approach is based on the Visual Analytics Mantra: First, we statistically analyze the data to identify those variables that separate DN patients from a control group. Afterwards, we show the important separating variables individually, but also in the context of all variables regarding a pre-defined classification. By doing so, we support the understanding of the categorization in respect of the value distribution of variables. This allows for zooming, filtering and further analysis like deleting non-relevant variables that do not contribute to the definition of markers as well as deleting data records with false data values or false classifications. Finally, outliers are observed and investigated in detail. So, a third group of potential DN patients can be introduced. In this way, the detection of early DN can be effectively supported.

In cell biology, models increasingly capture dynamics at different organizational levels. Therefore, new modeling languages are developed, e.g., like ML-Rules, that allow a compact and concise description of these models. However, the more complex models become the more important is an efficient execution of these models. t -leaping algorithms can speed up the execution of biochemical reaction models significantly by introducing acceptable inaccurate results. Whereas those approximate algorithms appear particularly promising to be applied to hierarchically structured models, the dynamic nested structures cause specific challenges. We present a t -leaping algorithm for ML-Rules which tackles these specific challenges and evaluate the efficiency and accuracy of this adapted t -leaping based on a recently developed visual analysis technique.

In this work, we aim at facilitating the analysis of spatial simulations of particles at the microscopic level. This level poses significant challenges to interactive visual analysis tools. On the one hand, the data may contain up to 100.000 data points, and on the other hand, the data exhibit Brownian motion. As a first step to deal with these challenges, we apply well-accepted techniques to visualize the data and to allow analysts to interact with the data and their visual representation. Preliminary results from a spatial simulation of protein-lipid-raft interaction indicate that interactive visual solutions are indeed a useful addition to the modeling and simulation toolbox.

In systems biology, analyzing simulation trajectories at multiple scales is a common approach when subtle, detailed behavior and fundamental, overall behavior of a modeled system are to be investigated at the same time. A variety of multiscale visualization techniques provide solutions to handle and depict data at different scales. Yet the mere existence of multiple scales does not necessarily imply the existence of additional information on each of them: Data on a more fine-grained scale may not always yield new details, but instead reflect the already known data from more coarse-grained scales – just at a higher resolution. Nevertheless, to be sure of this, all scales have to be explored. We address this issue by guiding the exploration of simulation trajectories according to information about the deviation of the data between subsequent scales. For this purpose, we apply different dissimilarity measures to the simulation data at subsequent scales to automatically discern heterogeneous regions that exhibit deviating behavior on more fine-grained scales. We mark these regions and display them alongside the actual data in a multiscale visualization. By doing so, our approach provides valuable visual cues on whether it is worthwhile to drill-down further into the multiscale data and if so, where additional information can be expected. Our approach is demonstrated by an exploratory walk-through of stochastic simulation results of a biochemical reaction network.

Visual Analytics offers various interesting methods to explore high dimensional data interactively. In this paper we investigate how it can be applied to support experimenters and developers of simulation software conducting simulation studies. In particular the usage and development of approximate simulation algorithms poses several practical problems, e.g., estimating the impact of algorithm parameters on accuracy or detecting faulty implementations. To address some of those problems, we present an approach that allows to relate configurations and accuracy visually and exploratory. The approach is evaluated by a brief case study, focusing on the accuracy of Stochastic Simulation Algorithms.

Smart environments integrate a multitude of different device assembles and aim to facilitate proactive assistance in multi-display scenarios. However, the integration of existing software, especially visualization systems, taking advance of this novel capabilities is still a challenging task. In this paper we present a smart view management concept for a integration that combines and displays views of different systems in smart meeting rooms. Considering varying requirements arising in such environments we provide a smart viewing management taking e.g. the dynamic user positions, view directions and even the semantics of views to be shown into account.

Visual analysis sessions are increasingly conducted in multidisplay environments. However, presenting a data set simultaneously on heterogenous displays to users is a challenging task. In this paper we propose a two-step mapping strategy to address this problem. The first mapping step applies primary mapping functions to generate the same basic layout for all output devices and adapts the object size based on the displaycharacteristic to guarantee the visibility of all elements. The second mapping step introduces additional visual cues to enhance the effectiveness of the visual encoding for different output devices. To demonstrate the Two-Step-Mapping we apply this concept to scatter plots presenting cluster data.

The use of transparencies is a common strategy in visual representations to guarantee the visibility of different overlapping graphical objects, especially, if no visibility-deciding order is given (e.g., importance, depth). Alpha-blending, however, could generate new colors that are not specified by the given color scale and overlapping shapes may become difficult to be separated visually and the selection of specific elements would be difficult. In this paper, we present a new approach for representing overlapping regions: Instead of blending different colors, our weaving technique separates the original colors and shapes are easier to differentiate. Due to a deterministic weaving order, all overlapping objects are visible. We apply our approach to scatter plot visualizations to enhance the communication of overlapping clusters.

The whole process of information visualization - from data acquisition to the final image - includes sources of uncertainties (e.g., during measuring, aggregation, sampling, mapping,...) that may significantly influence the assumptions about the data and the decisions made. Although there is no universal representation of uncertainties, a somehow common approach is to use fuzzy or blurred representations of that data. In this work, we use nonphotorealistic rendering (NPR) as it offers different parameterizable techniques to produce fuzzy representations - for the purpose of a more widespread uncertainty visualization. We demonstrate our approach by two exemplary NPR-techniques applied to information visualization to visualize uncertainties along with the related data.

The fastest existing labeling-algorithms allow the labeling of thousands of objects within a few milliseconds on today's desktop computers. Thus, it is possible to recalculate the labeling in dynamic scenes for every frame as it is demanded in interactive scenarios like information visualization. The main problem in such dynamic labeling environments is the lack of frame-to-frame coherence. Topology of label positions can change dramatically between consecutive frames – resulting in flickering and popping artifacts. Hence, visual label tracking becomes difficult and usability suffers. This short paper presents a universal approach for solving these problems by the use of animations – without manipulating the underlying labeling algorithm.

An important aim of information visualization is the communication of characteristics of the data. Beside the exploration of relevant aspects, presentation of the findings is crucial. Due to the increasingly large data volumes, however, new strategies to avoid cluttered displays are necessary. Our approach makes use of progressive refinement to deskew information temporally. Moreover, we also apply its beneficial properties to enhance the communication of data characteristics by a pre-defined Tour-through-the-data and to simplify the adaptation to different viewing devices.

In many information visualization techniques, labels are an essential part to communicate the visualized data. To preserve the expressiveness of the visual representation, a placed label should neither occlude other labels nor visual representatives (e.g., icons, lines) that communicate crucial information. Optimal, non-overlapping labeling is an NP-hard problem. Thus, only a few approaches achieve a fast non-overlapping labeling in highly interactive scenarios like information visualization. These approaches generally target the point-feature label placement (PFLP) problem, solving only label-label conflicts. This paper presents a new, fast, solid and flexible 2D labeling approach for the PFLP problem that additionally respects other visual elements and the visual extent of labeled features. The results (number of placed labels, processing time) of our particle-based method compare favorably to those of existing techniques. Although the esthetic quality of non-real-time approaches may not be achieved with our method, it complies with practical demands and thus supports the interactive exploration of information spaces. In contrast to the known adjacent techniques, the flexibility of our technique enables labeling of dense point clouds by the use of non-occluding distant labels. Our approach is independent of the underlying visualization technique, which enables us to demonstrate the application of our labeling method within different information visualization scenarios.

In many interactive scenarios, the fast recognition and localization of crucial information is very important to effectively perform a task. However, in information visualization the visualization of permanently growing large data volumes often leads to a simultaneously growing amount of presented graphical primitives. Besides the fundamental problem of limited screen space, the effective localization of single or multiple items of interest by a user becomes more and more difficult. Therefore, different approaches have been developed to emphasize those items – mainly by manipulating the items size, by suppressing the whole context or by adding supplemental visual elements (e.g., contours, arrows). This paper introduces the well known illustrative technique of haloing to information visualization to address the localization problem. Applying halos emphasizes items without a manipulation of size or an introduction of additional visual elements and reduces the context suppression to a locally defined region. This paper also presents the results of a first user-study to get an impression of the usefulness of halos for a faster recognition.

A not negligible number of information visualization techniques uses 3D-geometry to visualize data and structures. Thereby, constantly growing data volumes influence the final visual representation and often result in the occlusion of certain items. Therefore, different approaches have been developed that mainly manipulate item positions to uncover specific items of interest or otherwise use filtering and information hiding to reduce the amount of visible items. This paper presents a novel method to adapt 3D-views from information visualization by the use of the well-known illustrative technique ghost-view to successfully address this occlusion problem. Applying ghost-views to 3D information visualization techniques ensures the visibility of selected items by view-dependently manipulating the transparency of unselected data: without any manipulation of positions or continuous context suppression. Our approach is applicable to most 3D visualization techniques. It is interactive and easy to adapt to existing visualization environments.

Due to complexity, modern visualization techniques for large data volumes and complex interrelationships are difficult to understand for non-expert users and even for expert users the visualization result may be difficult to interpret. Often the limited screen space and the risk of occlusion hinders a meaningful explanation of techniques or datasets by additional visual elements. This paper presents a novel way how views from information visualization can be adapted by the use of the well-known illustrative technique exploded view, to successfully face the problems described above. The application of exploded views gains screen space for an explanation in a smart way and acts explanatory itself. With our approach of illustrating visual representations, the understanding of complex visualization techniques is eased and new comprehensible views on data are given.

Inspired by Venn diagram layouts, the Treemap is one of themost prevalent implicit tree visualization techniques. Ever since its publication, it has been modified and extended in many ways. This work presents a way to generate 3-dimensional Treemap visualizations by a 4-step procedure. It can be used for rapid prototyping and comparing different 3D Treemap layout approaches, to devise user studies on 3D Treemap layouts or for educational purposes.

Textual annotations are important elements in all but the most simple visual interfaces. In order to integrate textual annotations smoothly into the dynamic graphical content of interactive information systems, fast yet high-quality label layout algorithms are required. With the ongoing pervasion of mobile applications these requirements are shifted from workstations to comparatively low-performance mobile devices. Fortunately, ubiquitous network access is also on the advance, so that mobile applications can employ remote layout services on external workstations. This paper presents two novel label layout algorithms for relevancedriven dynamic visualizations in interactive information systems. They are employed to generate adaptive visualizations in a mobile maintenance support scenario.

Ca. 160 Millionen Computer haben heutzutage Zugang zum Internet – einem der größten, schnellstwachsenden und meistgenutzten Medien. Durch die ständig wachsende Struktur des Internets kommen allerdings nicht nur neue, sondern auch redundante und veraltete Daten in diesem Rechnerverbund hinzu und es stellt sich immer öfter die Frage, wo die relevanten Daten zu finden sind. Diese Arbeit untersucht deshalb einen verteilten suchunterstützenden Algorithmus, der vor allem auf der Relevanz von Daten beruht.