Robert S. Laramee 1 Visualization: An Introduction and Applications Robert S. Laramee Department of Computer Science Swansea University

Robert S. Laramee 2 Visualization: An Introduction and Applications Overview  Introduction  Visualization  Volume Visualization (+Application)  Flow Visualization (+Application)  Information Visualization (+Application)  Conclusions

Robert S. Laramee 3 Visualization: An Introduction and Applications Introduction: Who is Bob?  July 2006: Joined Computer Science Dept at Swansea University  : worked at VRVis Research Center (VRVis.at)– the bridge between academia and industry in Austria  2005: PhD, Computer Science, Vienna University of Technology (TUWien)  2000: Msc., Computer Science, University of New Hampshire, Durham, NH  1997: BSc., Physics, University of Massachusetts (ZooMass), Amherst, MA  Research in  Data visualization  Software Engineering  Human-computer interaction

Robert S. Laramee 4 Visualization: An Introduction and Applications Visualization: What is it? “The purpose of computing is insight, not numbers” [Richard W. Hamming, 1962] Visualization:  A tool that allows the user to gain insight into data.  To form a mental vision, image, or picture of (something not visible or present to the sight, or of an abstraction) ; to make visible to the mind or imagination [Oxford English Dictionary, 1989]  The non-fiction version of computer graphics

Robert S. Laramee 5 Visualization: An Introduction and Applications Visualization – Background  Visualization is very old  Often an intuitive step to make something clearer, e.g., a graph  Data set sizes are ever-increasing making a graphical approach necessary  Classical (easy) approaches known from business graphics (Excel, etc.)  Visualization = its own scientific discipline since ~1987  First visualization conference: 1990 L. da Vinci ( ) 1997:

Robert S. Laramee 6 Visualization: An Introduction and Applications User & Task  What is the problem?  User task/questions Visualization goals Examples:  surgery planning  filter analysis  stock market vis. [Oeltze et al., 2004] [Shneiderman/Wattenberg, 2001] [VRVis, 2004]

Robert S. Laramee 7 Visualization: An Introduction and Applications Three examples Data & Domain  Application domain  Data origin  Data organization  Data extent  Form of data [TIANI, 2001] [illustration by Kirby et al., 1999] [Bischi et al., 2004]

Robert S. Laramee 8 Visualization: An Introduction and Applications Typical Visualization Tasks Visualization is good for:  exploration  find the unknown, unexpected  hypothesis generation  analysis  confirm or reject hypotheses  information drill-down  presentation  communicate/disseminate results [Seo/Shneiderman 2004] [Doleisch et al., 2003] [Bruckner/ Gröller 2005]

Robert S. Laramee 9 Visualization: An Introduction and Applications Visualization: 3 Foci Three main sub-fields:  Volume Visualization  Flow Visualization  Information Visualization Scientific Visualization 3D nD Inherent geometry Usually no inherent geometry

Robert S. Laramee 10 Visualization: An Introduction and Applications From the Data Point of View Data with inherent location (and time) – “SciVis“  3D, 3D+time, 2D, etc.  scalar, vector, multi-variate, etc.  data organization: Cartesian grid, simulation grid, etc.  continuous Tabular data (lines  rows) – “InfoVis“  multi-variate data  heterogeneous data (nominal, categorical, etc.)  discontinuous (or discrete)

Robert S. Laramee 11 Visualization: An Introduction and Applications What is Volume Visualization? Volume Visualization is the depiction of Volume data  Picture/image 3D  2D  Projection (e.g., MIP), slicing, surface extraction, volume rendering, …  Volume Data =  3D  1D Data  scalar data, 3D data space, space filling (dense-as opposed to sparse)  User goals:  to gain insight into 3D Data  depends strongly on what user is interested in (focus + context)

Robert S. Laramee 12 Visualization: An Introduction and Applications About Volume Visualization Example a density volume from CT etc.  perceptual challenge (3D→2D)  selection (slices) vs. projection (rendering)  feature selection  performance challenge (billions of data items)  data structures (grids)  utilizing the GPU

Robert S. Laramee 13 Visualization: An Introduction and Applications Volume Visualization: An Application Medical Visualization

Robert S. Laramee 14 Visualization: An Introduction and Applications What is Flow Visualization?  a classic topic within scientific visualization  depiction of vector quantities (as opposed to scalar quantities)  applications include: automotive design, aerodynamics, astronomy, engineering, fluid mechanics, meteorology, oceanography, medicine, simulation, turbomachinery, Challenges: 1. to effectively visualize both magnitude + direction often simultaneously 2. large data sets 3. time-dependent data 4. What should be visualized? (data filtering/feature extraction)

Robert S. Laramee 15 Visualization: An Introduction and Applications What is Flow Visualization?  Challenge: to effectively visualize both magnitude + direction often simultaneously magnitude only direction only

Robert S. Laramee 16 Visualization: An Introduction and Applications  swirl motion: characterized by motion about cylinder-aligned axis  more stable (easier)  tumble motion: characterized by motion about axis orthogonal to cylinder  unstable, more difficult Flow Visualization: An Application

Robert S. Laramee 17 Visualization: An Introduction and Applications Achieving ideal patterns of motion leads to optimal mixing (of air and fuel) conditions  e.g., higher exhaust/gas ratio (EGR)  decrease in fuel consumption  lower emissions 1. Can visualization provide insight into or verify characteristic shape/behavior of flow? 2. What tools help to visualize swirl/tumble motion? 3. Where (in the combustion chamber) are ideal ideal flow pattern not being realized? Flow Visualization: An Application

Robert S. Laramee 18 Visualization: An Introduction and Applications Flow Visualization: An Application

Robert S. Laramee 19 Visualization: An Introduction and Applications InfoVis – Categorical Data Vis.  Example: set of returned questionaires (99 questions, most binary or categorical; ≈ lines) lots of columns many lines binary attribute, values just “0” or “1”

Robert S. Laramee 20 Visualization: An Introduction and Applications Project: Visualisation of Sensor Data from Animal Movement Edward Grundy Mark W. Jones Robert S. Laramee Rory P. Wilson Emily L.C. Shepard Visual Computing Group Institute of Environmental Sustainability

Robert S. Laramee 21 Visualization: An Introduction and Applications Background...to animals in the wild... To gather data on: acceleration, temperature, pressure, etc Biologists at Swansea have attached sensors... 21

Robert S. Laramee 22 Visualization: An Introduction and Applications 22 Old Visualization Technique 2D line plots of the acceleration data is difficult to interpret, Large time domain makes relating different periods difficult Three channels (possibly more) need to be correlated mentally by user Relating intensity plots to orientation or movement is difficult walkingwashingflying diving

Robert S. Laramee 23 Visualization: An Introduction and Applications Background Understanding animal behaviour helps us to understand the environment, but animals are difficult to follow (for good reason). We can follow them easily in some places... Or not at all... With more difficulty in others... 23

Robert S. Laramee 24 Visualization: An Introduction and Applications 24 Objectives Given accelerometry data, it is useful to: a)identify extraordinary events, a)identify similarity, a)relate to other attributes to form hypotheses a) identify commonly adopted postures While reducing cognitive effort required by line plots.

Robert S. Laramee 25 Visualization: An Introduction and Applications Acceleration data can be correlated with other attributes. Visualisation of Multiple Attributes Replacing the magnitude of acceleration vector with pressure attribute improves readability of visualisation. 25 Visualisation now achieves objective (c); relating accelerometry to other attributes to form a hypothesis for behaviour.

Robert S. Laramee 26 Visualization: An Introduction and Applications Visualisation of Distribution Frequently adopted orientations indicate common behaviours. These can be observed with a histogram over sphere; achieving objective (d) (identifying commonly adopted postures). 26

Robert S. Laramee 27 Visualization: An Introduction and Applications 27 Visualisation of Data Distribution This can be used to decide whether a point is “in- posture” or not. Allowing us to quantify the tortuosity of the transition between postures, and visualise this in a graph. Density based c-means clustering of the accelerometry produces a degree of membership value.

Robert S. Laramee 28 Visualization: An Introduction and Applications 28 Domain Expert Review Scatterplot and overlay are valuable tools for both exploration and communication of results. Open research problem in marine biology: “Why don’t diving birds get decompression sickness (the bends)?” Visual correlation of: Acceleration Pressure Resulted in hypothesis: “Diving birds slow their ascent to prevent nitrogen bubbles building up in their blood.” strata in ascent absent during descent

Robert S. Laramee 29 Visualization: An Introduction and Applications State Transition Diagrams Data clustering gives further insight into postures and energy expenditure. 29 Posture spheres present new interaction possibilities for statistical analysis.

Robert S. Laramee 30 Visualization: An Introduction and Applications Results: Animal Tracking Video

Robert S. Laramee 31 Visualization: An Introduction and Applications Summary  Driven by / driving the strongest human sense  Lots of different application fields  Serves different purposes  Exploration, analysis, presentation  Depends on  data characteristics, e.g., dimensionality  data model (continuous vs. discrete)  Rapidly developing research field

Robert S. Laramee 32 Visualization: An Introduction and Applications Acknowledgements  Thank you for your attention!  Any Questions? Thanks to the following people: Helmut Doleisch, Christoph Garth, Edward Grundy, Markus Hadwiger, Helwig Hauser, Mark W Jones, Robert Kosara, Lukas Mroz, Juergen Schneider, Emily Shepard, Xavier Tricoche, Rory P Wilson

Robert S. Laramee 33 Visualization: An Introduction and Applications Visualization Application: Virtual Reality