Introduction to visualization - Noderīgi resursi RTU DITF …ditf.afraid.org/ditf/2 kurss/Modeleeshanas un Imitaacijas... · · 2006-02-07© Arnis Lektauers, 2005 INTRODUCTION

IntroductionIntroduction to to VisualizationVisualization

©© Arnis Lektauers, 2005Arnis Lektauers, 2005

INTRODUCTIONINTRODUCTION

TOTO

VISUALIZATIONVISUALIZATION

Department of Modelling and Simulation

Riga Technical University



•• Visualization finds ancestry in pictogramsVisualization finds ancestry in pictograms– E.g. travel, Da Vinci’s airplanes, architecture – human

generated

History

Leonardo da Vinci's drawing of a helicopter

One of Leonardo da Vinci’s flying machines - the ornithopter.



History

•• ComputerComputer--generated since late 40generated since late 40’’ss– Large tables expressed as plots, statistical data for exploration

•• Mid 80Mid 80’’s: need and opportunity grows: need and opportunity grow– Increased data rates from

•• measuring devices: e.g. space missions, medical measuring devices: e.g. space missions, medical instrumentsinstruments

•• scientific computing: e.g. start of supercomputer scientific computing: e.g. start of supercomputer centers, computational sciences (CFD or Molecular centers, computational sciences (CFD or Molecular Modeling)Modeling)

– Mature and cheap technology: powerful graphical workstations, color, sufficient memory and storage



•• Committee on Committee on ““Graphics, Image Graphics, Image Processing, and WorkstationsProcessing, and Workstations”” (1986)(1986)– Sponsored by NSF / Division of Advanced Scientific

Computing

•• Goal of committeeGoal of committee– Recommendations to HW / SW builders to improve scientific

productivity

•• Result of committeeResult of committee– Recommendations to research communities to develop new

concepts / techniques for “Visualization in Scientific Computing (ViSC)”

History



•• ““To visualizeTo visualize””: form a mental vision, image, or picture of : form a mental vision, image, or picture of (something not visible or present to sight, or of an (something not visible or present to sight, or of an abstraction); to make visible to the mind or imagination abstraction); to make visible to the mind or imagination [The Oxford English Dictionary, 1989][The Oxford English Dictionary, 1989]

•• ““VisualizationVisualization””: a computer generated image or collection : a computer generated image or collection of images, possibly ordered, using a computer of images, possibly ordered, using a computer representation of data as its primary source and a human representation of data as its primary source and a human as its primary target.as its primary target.

•• ComputerComputer--generated visualizations meant to be viewed by a generated visualizations meant to be viewed by a human includes various flavors of visualization:human includes various flavors of visualization:

•• Data VisualizationData Visualization•• Scientific VisualizationScientific Visualization•• Information VisualizationInformation Visualization•• Software VisualizationSoftware Visualization

Definitions



Definitions

•• Mapping from computer representations to perceptual Mapping from computer representations to perceptual (visual) representations, choosing encoding techniques to (visual) representations, choosing encoding techniques to maximize human understanding and communicationmaximize human understanding and communication

Mapping Process

DomainNumbers / Data

DomainPicture(s)

RealityComputer

representationof reality

Picture(s) Viewers



Visualization and adjacent Disciplines

•• Computer GraphicsComputer Graphics– Efficiency of algorithms (CG) versus effectiveness of

use (V)

•• Computer VisionComputer Vision– Mapping from pictures to abstract description (CV)

versus mapping from abstract description to picture domain (V)

•• Image ProcessingImage Processing– Mapping from data domain to data domain (IP) versus

mapping from data domain to picture domain (V)



Visualization and adjacent Disciplines

•• (Visual) Perception (Visual) Perception – General and scientific explanation of human abilities

and limitations (VP) versus goal oriented use of visual perception in complex information representation.

•• Art and DesignArt and Design– Aesthetics and style (AD) versus expressiveness and

effectiveness of data (V)



A Model of Perceptual Processing

A three-stage model of human visual information processing



Examples of Goals in Visualization

•• Exploration / exploitation of data and informationExploration / exploitation of data and information•• Enhancing understanding of concepts and Enhancing understanding of concepts and

processesprocesses•• Gaining new (unexpected, profound) insightsGaining new (unexpected, profound) insights•• Making invisible visibleMaking invisible visible•• Effective presentation of significant featuresEffective presentation of significant features•• Quality control of simulations, measurementsQuality control of simulations, measurements•• Increasing scientific productivityIncreasing scientific productivity•• Medium of communication / collaborationMedium of communication / collaboration



Sample Applications

•• GeophysicsGeophysics, e.g. , e.g. ““The Visualization of StormThe Visualization of Storm””•• BiochemistryBiochemistry, e.g. the visualization of DNA, molecules, or , e.g. the visualization of DNA, molecules, or

crystalscrystals•• Engineering and PhysicsEngineering and Physics, e.g. the visualization of a helicopter , e.g. the visualization of a helicopter

turbine, of a wind tunnel, of the Big Bang, of Finite Elements turbine, of a wind tunnel, of the Big Bang, of Finite Elements Analysis computationsAnalysis computations

•• Sociology and PoliticsSociology and Politics, e.g. the visualization of census data, of , e.g. the visualization of census data, of vote distributions or the spread of aidsvote distributions or the spread of aids

•• MathematicsMathematics, e.g. the visualization of , e.g. the visualization of splinessplines•• Information TechnologyInformation Technology, e.g. the visualization of the web, the , e.g. the visualization of the web, the

visualization of retrieved documents from queryvisualization of retrieved documents from query



Examples of Visualization

This graphic is an adaptation of M. Charles Joseph This graphic is an adaptation of M. Charles Joseph MinardMinard’’ss ““March of the Napoleon March of the Napoleon ArmyArmy”” by Sunny McClendon, as part of an Information Design Class at tby Sunny McClendon, as part of an Information Design Class at the University he University of Texas at Austinof Texas at Austin




““Study of a Numerically Modeled Severe StormStudy of a Numerically Modeled Severe Storm””, Video by , Video by WilhelmsonWilhelmson, Robert et al. , Robert et al. (Department of Atmospheric Studies and NCSA)(Department of Atmospheric Studies and NCSA)




Simulation of Shuttle debris trajectory. A 1.67 pound slab of inSimulation of Shuttle debris trajectory. A 1.67 pound slab of insulating foam is seen sulating foam is seen falling off the external tank after falling off the external tank after Columbia'sColumbia's launch and hitting the left wing. The launch and hitting the left wing. The Columbia Accident Investigation Board (CAIB) has identified a deColumbia Accident Investigation Board (CAIB) has identified a debris event like this bris event like this as the most likely cause of the as the most likely cause of the ColumbiaColumbia disaster. This image was used in the disaster. This image was used in the CAIB'sCAIB'sfinal report. (NASA Advanced Supercomputing Division)final report. (NASA Advanced Supercomputing Division)




External Truck Aerodynamics Animation. Models such as this will lead to modifications that increase fuel efficiency by minimizing drag force. Reducing drag force will also lessen environmental pollution and improve stability and vehicle control. (Old Dominion University)




Simulation and Animation Technology for Addressing Environmental Homeland Security Concerns (HidroGeoLogic, Inc.)



Impact of Future Technology

•• Next generation PCsNext generation PCs•• Next generation storage systemsNext generation storage systems•• Next generation display technologiesNext generation display technologies•• Next generation communication systemsNext generation communication systems•• Next generation analytic toolsNext generation analytic tools•• Limitation of human capacityLimitation of human capacity•• Improved understanding of psychological and Improved understanding of psychological and

perceptual issuesperceptual issues



The Data

•• The data := The data := ““data generated from mathematical data generated from mathematical models or computations and from human and models or computations and from human and machine collectionmachine collection””

Data and the world being modeledData and the world being modeled– Establish valid and reliable relationship between data

and world– Scientific objectives and method

•• Attempt to explain the real worldAttempt to explain the real world•• UnderstandingUnderstanding•• PredictionPrediction•• Create a model of the worldCreate a model of the world•• Acquire data to verify or refine the modelAcquire data to verify or refine the model

Definition and Goals



The Data

•• Data model = conceptual view of dataData model = conceptual view of dataCharacterize data by e.g.Characterize data by e.g.

–– GeometryGeometry–– TopologyTopology–– ValueValue

•• Advantage of data modelsAdvantage of data models– Discipline independent view of data– Choose expressive visualization technique– Avoid “mental road blocks”

Definition and Goals



The Data

Overview of selected data characteristics

•• Nominal, ordinal, quantitativeNominal, ordinal, quantitative•• Point, scalar, vectorPoint, scalar, vector•• ““continuouscontinuous”” datadata•• Topology / structure for nonTopology / structure for non--continouscontinous

datadata•• Data reliabilityData reliability•• Valid range of dataValid range of data•• Time descriptorsTime descriptors



The Data

Nominal, ordinal, quantitative

•• Nominal data Nominal data –– members of certain class, e.g. members of certain class, e.g. [Riga, [Riga, TalsiTalsi, , VentspilsVentspils, , LiepajaLiepaja], or [Maple, Birch, ], or [Maple, Birch, Oak]Oak]– Effective visual attributes: color (hue), symbol

•• Ordinal data Ordinal data –– related by order, e.g. [low, medium, related by order, e.g. [low, medium, high], or [tiny, small, medium, large]high], or [tiny, small, medium, large]– Effective visual attributes: brightness, size, (color – hue, if

meaningful to observer)

•• Quantitative data Quantitative data –– carry precise numerical value, carry precise numerical value, e.g. [2.3, 4.56, 0.8, 2.5Ee.g. [2.3, 4.56, 0.8, 2.5E--35]35]– Effective visual attributes: position, length



Point, Scalar, Vector

The Data

Syntactical categories, additionally characterized by dimensionsSyntactical categories, additionally characterized by dimensions

•• PointPoint– Each data element is considered a position in n-dimensional

space– Example: measurements of leaves: [length, width, tree type,

age], e.g. [2.3, 1.2, B, 1], [4.3, 2.2, B, 3], [1.5, 1.5, M, 1], [3.0, 2.9, M, 3], …

– Expressive visualizations: scatter plots, glyphs

•• ScalarsScalars– Each data element has a numeric expression– Example: topography of terrain, expressed as 2-d field

containing elevations



The Data

Point, Scalar, Vector

•• Scalar arrays Scalar arrays –– often often ““discrete samples of discrete samples of continuous functionscontinuous functions””– Usually 1 (linear), 2 (image), or 3 (volumetric) dimensional data

sets; samples in equidistant or non-equidistant steps– Expressive visualizations: line graph, shaded surface, volume

viewing

•• VectorsVectors– Each data element is considered a straight directed line with a

certain length (magnitude) in n-dimensional space– Examples: direction of particle flow in channel– Expressive visualizations: arrows, stream lines, particle tracks



The Data

“Continuous” Data

““ContinuousContinuous”” data can be represented by (samples of)data can be represented by (samples of)function: function: yiyi = = fifi (X), where X = (x1 , x2 , x3 , ..., (X), where X = (x1 , x2 , x3 , ..., xnxn ); i=[1,...,m]); i=[1,...,m]

x .... independent variables; x .... independent variables; e.ge.g space, time, spectral (space, time, spectral (““dimensionsdimensions””))y .... dependent variables (y .... dependent variables (““parametersparameters””))

Comes in regular and irregular formatsComes in regular and irregular formats

Expressive visualizations of functions: Expressive visualizations of functions: similar to scalar, similar to scalar, quantitative, ordinalquantitative, ordinal

Interpolation methods: Interpolation methods: must be meaningful in problem spacemust be meaningful in problem space

Computation time Computation time for visualization techniques faster on regular gridsfor visualization techniques faster on regular grids



Topology / structure of non-continous data

The Data

• Types of topology/structure, e.g.– Sequential (text)– Hierarchical– Relational– Single points and connectors

• Examples and corresponding expressive visualizations– Molecules (e.g. ball-and-stick model)– Data bases (cone tree; perspective wall)



The Data

Other data characteristics in a data model

• Data reliability– Missing data or unreliable data

• expressive visualizations: error bars; indicate borders between• real/missing data• careful with interpolation

• Valid range of data– min / max / mean / median

• Time descriptors– Various meanings of time: simulation time, simulated/actual

time frame, computation time, recording and playback time, user's time frame

– “time models” to support time conversions necessary to synchronize



The Data

Examples

Perspective Wall and Cone Tree: from CACM April 1993, InformationVisualizer by Robertson, Stuart and Mackinlay.

• Complex data sets and their visual counterparts, e.g- scientific visualization - proteins- software - web pages



Mapping

Visual Data Mapping

•• Need for systematic strategies (concepts, methodologies, Need for systematic strategies (concepts, methodologies, intelligent visualization systems) to exploit dataintelligent visualization systems) to exploit data

Hardware / Software User: abilities,disabilities, desires

+Visualization Goal

Numbers / Data

Graphical objects+

Visual attributesPicture(s)



Representation

Histograms (1-d and 2-d), Pie and Bar Charts

• Representative data characteristics– 1-d arrays of scalars, continuous or discrete data values

•• TechniquesTechniques– Bar chart: length of bar indicates value of (class of) items– 1-d histogram: length of bar indicates number of elements in

sub-category– 2-d histogram: brightness/color indicates number of elements

in sub-category– Pie chart: sector of circle indicates values of (class of) items

as fractions of a whole



Representation

Examples of Histograms, Pie and Bar Charts



Representation

Line Graphs

• Representative data characteristics– 1-d, (continuous), scalar data arrays, e.g. y = f(x)

• Technique– Curve drawn through single data points

• Special note on effectiveness– No mental interpolation necessary– Use interpolation method meaningful to problem space



Representation

Examples of Line Graphs



Representation

(n-dimensional) Scatter Plots

• Data characteristics: multivariate data space, such as botanical observationsTechnique– Define coordinate system appropriate for data– Project data and coordinate system to display space– Use points or symbols to define data element locations

• Effectiveness– Position is primary visual cue– Animation (change of view point) for 3-d effect– Dimensions > 2: use projections (“Grand Tour”)

Interaction: control over view point, rotation, “rocking”; “conditional box”



Representation

Example of 4-d Scatter Plot



Representation

Glyphs / Icons

• Representative data characteristics– Multivariate data spaces, such as computer performance

measurements, census data

• Technique– Define 1, 2, or 3 data variables as spatial dimensions– Compose small graph (glyph/icon) for each additional variable– Display each glyph as “complex pixel” in 1,2,or 3d space

• Special note on effectiveness– Distinguish between macroscopic/microscopic interpretation

of glyphs– Several visual attributes used in each glyph– “The whole is greater than the sum of its parts” (Gestalt

Theory)



Chernoff faces. Different data variables are mapped to the sizes and shapes of different facial features.

Representation

Example of Glyphs



Representation

Contour Lines (Isolines)

• Representative data characteristics– 2-d scalar data arrays, e.g. z = f(x,y), such as elevation map

• Technique– Trace lines of constant value (= threshold value) of 2-d raster

• Special note on effectiveness– Annotate selected isolines



Representation

Examples of Contour Plots



Representation

Examples of Contour Plots



Surface View

Representation

• Representative data characteristics– 2-d scalar data arrays, e.g. z = f(x,y), such as elevation map

• Wireframe technique– Treat “z” as elevation over 2-d terrain and use projection from

3-d to 2-d– Project mesh of lines parallel to x and y axes

• Shaded surface technique– Treat “z” as elevation over 2-d terrain and use projection from

3-d to 2-d– Project each data element / remove hidden surfaces– Assign grey value / color value

• Special note on effectiveness– Source of grey value/color value must be transparent to viewer



Example of Surface Plot



Image Display

Representation

• Representative data characteristics– 2-d scalar data arrays, e.g. z = f(x,y), such as LANDSAT image

• Technique– Straightforward: map each 2-d data element to brightness or

color of screen pixel

• Special note on effectiveness– Color / brightness scale necessary



Representation



Representation

Color Transformations

• Representative data characteristics– up to three scalar data arrays defined over same two

dimensions, e.g. zi = f(x,y), i=1,2,3 such as three TM (Thematic Mapper) channels of same terrain

• Technique– choose same technique (e.g. image display or surface) for

each data array– read zk = f(i,j), k=1,2,3 for each pixel location on screen,

resulting in 3 brightness values (z1, z2, z3)– use (z1, z2, z3) as coordinates to color space, e.g. RGB, HSV

\xdf ‘color’– use ‘color’ to paint pixel at screen location (i, j)



Representation

Examples of Color Transformations

Digital Elevation Map of Oetztal, Austria: hue is elevation; intensity is illumination

RGB transformation of three IRAS imagesData by NASA/JPL



Representation

Volume Slices

• Representative data characteristics– 3-d scalar data arrays, e.g. w = f(x,y,z), such as medical scans

of human organs

• Technique– intersect 2-d plane(s) with volume– use image display for visual representation– project planes to screen

• Special note on effectiveness– use appropriate coordinate system to depict location of

plane(s) in volume– animation (change of view point), hidden surfaces and

perspective geometry for 3-d effect



Representation

Examples of Volume Slices

Specimen from Visible Human Male –Head subset

Specimen from Visible Human Male –Thorax subset



Representation

Surface Rendering

• Representative data characteristics– 3-d scalar data arrays, e.g. samples of w = f(x,y,z), where w

(voxel value) might indicate color, opacity, density, material, or time

• Technique– surface reconstruction (define surfaces in 3-d raster) (e.g. by

using marching cubes algorithm or surface detection)– surface rendering (illumination, shading, projection)



Representation

Volume Viewing

• Representative data characteristics– 3-d scalar data arrays

• • Technique– Project volumetric data elements onto the display space, by

either– Backward projection (object-order): scan voxel space and

project to screen– Forward projection (image-order): scan screen pixels and

determine voxel contributions– Combination– Assign pixel brightness/color

• Special note on effectiveness– Transparency / translucency



Representation

Example of Volume Rendering



Example of Direct Volume Rendering

Materials Science: microstructures



Representation

Arrows

• Representative data characteristics– Vector fields

• Technique– Use arrow as glyph, vary following attributes of arrow

depending on variables: direction/length/width/reflection properties of shaft, type/color of arrow head

• Special note on effectiveness– Avoid cluttering by reducing amount of data to display– Additional problems in 3-d through directional ambiguity



Examples of Vector Plots

Representation



Interaction Issues

General Considerations

• Interaction := user’s actions during the visualization process

• Various forms of interaction– change of visuals: color, view point, shading techniques– change of visualization techniques– change of parameters for data generation (interactive steering)– exploration of data: explore data values; remove occluding

data



Interaction Issues

Human Performance Limitations

• users prefer shorter response times• feedback to user in less than 0.1 sec for continuous user input• response times > 15 sec are disruptive• shorter response times leads to shorter user think time• faster pace may increase productivity, but may increase error

rates• response time should be appropriate to task (changes from

complex task to atomic actions)• empirical tests can help to set suitable response times

Documents

Introduction to visualization - Noderīgi resursi RTU DITF …ditf.afraid.org/ditf/2 kurss/Modeleeshanas un Imitaacijas... · · 2006-02-07© Arnis Lektauers, 2005 INTRODUCTION