Visualizing Real-Time Multivariate Data Using Preattentive ...

1 Visualizing Real-Time Multivariate data UsingPreattentive ProcessingCHRISTOPHER G. HEALEY, KELLOGG S. BOOTH, and JAMES T. ENNSThe University of British ColumbiaA new method is presented for Visualizing data as they are generated from Real-Time ap-plications. These techniques allow viewers to perform simple data analysis tasks such asdetection of data groups and boundaries, target detection, and estimation. The goal is todo this rapidly and accurately on a dynamic sequence of data frames. Our techniques takeadvantage of an ability of the human visual system called Preattentive processing. Preat-tentive processing refers to an initial organization of the visual system based on operationsbelieved to be rapid, automatic, and spatially parallel.

2 Examples of visual features that canbe detected in this way include hue, orientation, intensity, size, curvature, and line believe that studies from Preattentive processing should be used to assist in the designof visualization tools, especially those for which high speed target, boundary, and regiondetection are important. Previous work has shown that results from research in preattentiveprocessing can be used to build visualization tools which allow rapid and accurate anal-ysis of individual, static data frames. We extend these techniques to a dynamic real-timeenvironment. This allows users to perform similar tasks on dynamic sequences of frames,exactly like those generated by Real-Time systems such as visual interactive simulation .

3 Westudied two known Preattentive features, hue and curvature. The primary question investi-gated was whether rapid and accurate target and boundary detection in dynamic sequencesis possible Using these features. Behavioral experiments were run that simulated displaysfrom our Preattentive visualization tools. Analysis of the results of the experiments showedthat rapid and accurate target and boundary detection is possible with both hue and curva-ture. A second question, whether interactions occur between the two features in a real-timeenvironment, was answered positively. This suggests that these and perhaps other visualfeatures can be used to create visualization tools that allow high-speed multidimensionaldata analysis for use in Real-Time applications.

4 It also shows that care must be taken inthe assignment of data elements to Preattentive features to avoid creating certain visualinterference and Subject Descriptors: [Information Interfaces and Presentation]:This work was supported in part by the National Science and Engineering Research Council ofCanada under various operating grants, by the British Columbia Advanced Systems Institute, andby the University and Province of British addresses: Healey and Booth, Imager Computer Graphics Laboratory, Depart-ment of Computer Science, 2366 Main Mall, University of British Columbia, Vancouver, BritishColumbia, V6T 1Z4, Canada; James T. Enns, Department ofPsychology, 2136 Main Mall, University of British Columbia, Vancouver, British Columbia, V6T1Z4, Canada; Permission to make digital/hard copy of part or all of thiswork for personal or classroom use is granted without fee provided that the copies are not made ordistributed for profit or commercial advantage, the copyright notice, the title of the publication, andits date appear, and notice is given that copying is by permission of the ACM, Inc.

5 To copy otherwise,to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or 1995 ACM 1049-3301/95/0700-0190 $ Interfaces ergonomics, screen design (graphics, colour); [Computer Graph-ics]: Methodology and Techniques ergonomics, interaction techniques; [Simulationand Modeling]: simulation Output AnalysisGeneral Terms: Design, Experimentation, PerformanceAdditional Key Words and Phrases: boundary detection, cognitive psychology, curvature,icon, hue, human vision, Multivariate data , Preattentive , scientific visualization, target de-tection, visual interactive simulation1. INTRODUCTIONThe field of scientific visualization draws on research from a wide spectrum oftraditional disciplines.

6 These include computer science, psychology, and the visualarts. The domain of visualization , as defined by a National Science Foundationpanel on scientific computing, includes the development of specific applications,the development of general purpose tools, and the study of research problems thatarise in the process [McCormick et al., 1987; Rosenblum, 1994]. To date, mostresearch efforts have focused on ad hoc visualization applications. Relatively fewefforts have formulated general guidelines for the design of visualization this paper, we report on new work that derives from an area of cognitive psy-chology known as Preattentive processing. This work is part of an on-going in-vestigation whose goal is a set of guidelines for visualization design.

7 We examinevisualization techniques for dynamic sequences of multidimensional data frames,like those produced by visual interactive simulation or Real-Time systems. We firstdefine a set of visualization requirements that are common to these then review research in Preattentive processing in order to establish abilitiesand limitations of human vision that are relevant to these requirements, after whichwe describe a scientific visualization tool we have developed that is based on thesegeneral considerations. Finally, we discuss the implications of our approach, bothfor specific applications and for the development of general guidelines in VISUALIZATION REQUIREMENTSS cientific visualization as a discipline within computer graphics is a relatively re-cent development.

8 The first reference to scientific visualization per seoccurredsometime in the late 1980s, although as [Fournier, 1994] has pointed out, manyaspects of scientific visualization have long been part of computer graphics. Pan-els and workshops in a variety of different disciplines are now addressing scientificvisualization and its relationship to their work [Becker and Cleveland, 1991; Rosen-blum, 1994; Treinish et al., 1989; Wolfe and Franzel, 1988]. The area is expandinginto a number of subfields that use computer graphics to solve various types ofproblems. Examples include volume visualization, medical imaging, flow visual-ization, and Multivariate data visualization.

9 Current research is concerned with theACM TOMACS 5, 3, July 1995, 190 221 HEALEY,BOOTH,ENNS design of intelligent visualization tools. Scientists are turning to computer graphics,psychology, and visual arts to understand how the human visual system analysesimages. This has led to the use of specific visual properties to make displays moreintuitive. These visual properties take advantage of the fundamental workings ofthe visual system applications require techniques for displaying data in Real-Time . One ex-ample is air traffic control, where displays are often shared by different operators,who acquire visual data from different parts of the display at the same time . Thevisualization technique must allow a variety of tasks to be performed rapidly andaccurately on dynamically changing subsets of the overall display.

10 Medical imag-ing systems such as CT, MRI, PET, and ultrasound are another type of applicationthat could benefit from Real-Time visualization. A method that allowed rapid andaccurate visual analysis of more than one aspect of the data would decrease theamount of time needed to complete the diagnostic task. This is important, becausethese types of systems often cannot be time -shared by multiple users. Any improve-ment in visualization would increase total throughput for the system. Moreover,better displays might reduce errors made during analysis. Even a small increase inaccuracy is important in this type of simulation Visualization SystemsThe requirements for Real-Time applications are similar to another class of problems,the visualization of output from simulation systems.