The study of fluid behaviors has been a challenging topic. Flow visualization enables us to visually acquire qualitative and quantitative flow information. There exist various software tools performing different flow visualization tasks. However, we lack tools that help students learn important flow field concepts. In this paper, we present a visualization app, named FlowVisual which runs on iOS devices, to illustrate basic flow field concepts in 3D. In order to meet a comprehensive learning goal for students, we integrate a number of techniques into FlowVisual design, including field-line tracing, field-line comparison, critical point detection and classification, template-based seeding, and surface visualization. We evaluate and demonstrate the effectiveness of FlowVisual by conducting a formal user study including an introduction and training session, an auto-grading test, and a post-questionnaire survey.

Morse decomposition has been shown a reliable way to compute and represent vector field topology. Its computation first converts the original vector field into a directed graph representation, so that flow recurrent dynamics (i. e., Morse sets) can be identified as some strongly connected components of the graph. In this paper, we present a framework that enables the user to efficiently compute Morse decompositions of 3D piecewise linear vector fields defined on regular grids. Specifically, we extend the 2D adaptive edge sampling technique to 3D for the outer approximation computation of the image of any 3D cell for the construction of the directed graph. To achieve finer decomposition, a hierarchical refinement framework is applied to procedurally increase the integration steps and subdivide the underlying grids that contain certain Morse sets. To improve the computational performance, we implement our Morse decomposition framework using CUDA. We have applied our framework to a number of analytic and real-world 3D steady vector fields to demonstrate its utility.

In this paper, we apply a derived scalar field, called the φ field, to assist the visualization of various flow data. The value of the φ field at a spatio-temporal location is determined by the accumulated angle changes of the tangent directions along the integral curve starting from this location. Important properties of the φ field and its gradient magnitude |∇φ| field are studied. In particular, we show that the patterns in the derived φ field are generally aligned with the flow direction based on an inequality property. In addition, we compare the φ field with some other attribute fields and discuss its relation with a number of flow features, such as topology, LCS and cusp-like seeding structures. Furthermore, we introduce a unified framework for the computation of the φ field and its gradient field, and employ them to a number of flow visualization and exploration tasks, including integral curve filtering, seeds generation and flow domain segmentation. We show that these tasks can be conducted more efficiently based on the information encoded in the φ field and its gradient.

Image local information is crucial for accurate segmentation of images with inhomogeneities in the intensity. In many popular methods, however, local regional information is normally underestimated and not included in the segmentation framework. In this paper, a segmentation is formalised as Bayesian Inference procedure. By considering the spatial variation of the intensity distribution, a likelihood that contains a joint distribution of intensity and spatial location is obtained. After incorporating this likelihood in a Bayesian maximum a posteriori estimation, we transformed the stochastic model into a segmentation method which utilises local information to segment images with inhomogeneities so as to guarantee a global optimum for the two-region case(foreground/background). Taking the computational complexity of our model into account, we take advantage of a GPU parallel algorithm to accelerate computation without losing segmentation accuracy. We demonstrate that taking into account local image information, our method results in significant improvements for image segmentation.

Understanding how brain regions are interconnected is an important topic within the domain of neuroimaging. Advances in non-invasive technologies enable larger and more detailed images to be collected more quickly than ever before. These data contribute to create what is usually referred to as a connectome, that is, a comprehensive map of neural connections. The availability of connectome data allows for more interesting questions to be asked and more complex analyses to be conducted. In this paper we present a novel web-based 3D visual analytics tool that allows user to interactively explore the intrinsic geometry of the connectome. That is, brain data that has been transformed through a dimensionality reduction step, such as multidimensional scaling (MDS), Isomap, or t-distributed stochastic neighbor embedding (t-SNE) techniques. We evaluate our tool through a series of real-world case studies, demonstrating its effectiveness in aiding domain experts for a range of neuroimaging tasks.

In this article the authors investigated a visualization tool (uVis) for end-user developers, in order to see how end users actually use it. The tool was an early version and the investigation helped the authors to improve it. The investigation showed that users appreciated the simple formula language, the coordinated panels, and the drag-and-drop mechanism. However, the most important thing for them was the immediate response when they changed something, for instance part of a formula. The entire visualization was updated immediately without having to switch from development view to production view. With uVis, developers construct a visualization from simple visual components such as boxes, curvePoints, and textboxes. All component properties such as Top and BackColor can be complex formulas similar to spreadsheet formulas. The operands in the formula can address relational data in a database, other visual objects, and dialog data provided by the user. A special Rows property can bind to a database query and make the component replicate itself for each row in the query. In this way, traditional as well as novel visualizations can be constructed. The most serious usability problems were data binding and not noticing errors (errors were shown in an error list, but not in the formula that had the error). There were many other usability problems. Removing them would speed up learning and make the tool more successful. © 2016 Society for Imaging Science and Technology.

Interactive highlighting is a common component of many techniques used in visualization such as brushing and dynamic queries. Sometimes these may be used in combination necessitating that two different highlighting methods be simultaneously applied. The challenge of effective highlighting is to design methods that make a subset of the items on a display stand out clearly without overly interfering with other information on a display. This is especially difficult when more than one subset of displayed symbols must be simultaneously highlighted. Three experiments are reported that investigate four different highlighting methods: 3D vs 2D symbols, encirclement, oscillatory motion and blinking. These are applied to the nodes in node-link diagrams. The first experiment was designed to evaluate the highlighting methods used separately and the results showed all four techniques to be effective. The second experiment evaluated combinations of highlighting methods. E.g. can we easily find a node that is both moving AND 3D in a set of nodes some of which are 3D and some of which are moving. The results showed that combinations including motion were the most effective. The third experiment was designed to determine which highlighting methods, used both separately and in combination supported the rapid counting of small numbers of targets. Again, combinations using motion were the most effective.

When analysing data and handling a visualisation, users mainly spend their cognitive resources making sense of the graphical representation and mapping it back to the data and domain. This task becomes even more critical when dealing with larger data sets. Therefore, a valuable visualisation design strategy is to couple graphical representations and user tasks to better support the sense making process. This paper focuses on a particular task where users must make sense of state changes occurring on nodes of a graph. To this end, we propose JASPER, a new layout algorithm focusing on the visualisation of nodes inspired from pixel-oriented layouts, relying on node clustering to identify and represent existing connections through spatial adjacency. JASPER can layout moderate size graphs in real-time and is able to tackle large graphs with up to 2 million nodes and 5 million edges in reasonable time (about half a minute). Furthermore, although JASPER has been designed around a specific application, the underlying methodology can be employed to draw quick overviews of any type of graphs. The paper lays down the underlying principles of JASPER, and reports it performances (execution times) on increasingly large graphs. JASPER is then used and showcased to visualise network propagation phenomenon in large graphs.

Interactive visualizations of complex datasets are an important tool for data exploration, but finding relationships between variables in highly multivariate environments often requires domain-tailored combinations of visualization techniques. We address this challenge in the domain of air traffic flow analysis with a tool designed to show how days with air traffic management initiatives are influenced by the weather on those days and to explore how various day cluster analyses may provide insight into relationships between the measured weather events and air traffic management. Our tool, called Weatherbin, provides both a broad overview of the day clusters as well as a detailed view of the weather conditions on any individual day, with interactive features to connect day details back to overall cluster averages.