Computer Graphics and Visualization in a Computational Science - - PowerPoint PPT Presentation

Computer Graphics and Visualization in a Computational Science Program

Steve Cunningham

California State University Stanislaus

Oregon State University, October 16, 2000

The imperative to scientific visualization comes from two sources: The need to understand sophisticated ideas more deeply The need to use sophisticated computation to do science

Visualization uses visual thinking: a key tool to help the student develop his or her deeper understanding of science

Visualization involves computation, the third leg of the tripod that supports science education and practice

• Theory
• Laboratory
• Computation

Scientific visualization is where computation and visual thinking meet - it’s using computation to support the visual understanding

• f science

How do we introduce students to visual thinking and computation in science?

• User level approach

– Tutorial software – Generalized tools

• Programming level approach

– Focus on creating specific visual content

These may not be equally good

• Tutorial software

– There seems to be an understanding that seeing visualizations provides little learning that texts cannot provide; the key is doing modeling and visualization themselves

• Generalized tools

– This approach is only as good as the tools used, and student learning may not transfer when the tools are replaced with new paradigms

Value of programming approach

• The student creates the linkage between the

image and the simulation or experimental data, and has many more ways to control the visual communication in the image

– Color set – Rendering options – Animation – Geometry

This approach can be embodied in a computer graphics course designed to be a core component

• f the student’s computational

science background

Traditional wisdom says that this might not be worthwhile:

• Computer graphics is thought to be a difficult

subject

• Computer graphics is thought to require a

student to to master highly technical algorithms

• Computer graphics is thought to be just about

making realistic images

BUT...

• Computer graphics does do not need to be

an especially difficult subject

• Computer graphics courses does not need

to require a focus on technical algorithms

• Computer graphics courses can focus on

visual communication and problem solving in application areas instead of simulating realism

My goal: to create a model for a computer graphics course that serves a broad student audience and is still a sound computer science course

Whom will the course serve?

• Shift the emphasis from developing

graphics specialists to developing a broad group of students with graphics skills

• Students can come from the sciences or

from many other disciplines, depending on the focus of the institution

• Computational science is a natural!

What is the new course model like?

• Focus is on graphics programming instead of

graphics theory, algorithms, and techniques

• Emphasizes visual thinking and communication
• Uses a standard programming API, such as

OpenGL, for its work

• Lectures discuss graphics concepts, while the

course projects allow the students to work in their individual specialty areas such as the sciences

What background is needed?

• Sound programming skills, and an ability

to see the geometry in their field

– Programming skills means roughly B or better in two programming courses – Seeing geometry requires simple spatial abilities that don’t come from coursework but can be picked up from the students’ work in their fields, especially science

Course projects

• Graphics topics, in order:

– Simple geometry and color – Lighting/shading, transformations, callbacks – Event-driven programming, user control, interface – Clipping, transparency, texture maps, splines, ... – Object selection and interaction with image

• Include problem statement as project source

and problem summary with project results

So what we have is not the traditional computer graphics course content of

Geometry Display

Rendering

but rather the more complete

Geometry Display

Data/Simulation

Geometrizing Rendering

Information & Insight

Possible student projects

Who wins with this approach?

• Computer science wins because we serve
• ur colleagues and our universities better
• Science students win because they get a

good background in the computer graphics they will use for their professional work

• Computer science students win because

they get useful professional skills

A key question that I have not yet worked on:

• How could we adapt some of the ideas, and

some of the content, from this course into a module that could take approximately four weeks of class and could be integrated into a computational science course on modeling and simulation, or into a more general scientific visualization course?

Contact information and credits:

• Email address is rsc@cs.csustan.edu or

cunningham@siggraph.org

• Draft notes and other materials are online at

http://www.cs.csustan.edu/~rsc/NSF /

• Thanks to Mike Bailey of SDSC for

valuable help and collaboration

This work is supported by National Science Foundation grant DUE-9950121. All opinions, findings, conclusions, and recommendations in this work are those of the author and do not necessarily reflect the views of the National Science Foundation.