Home | Lehre | Videos | Texte | Vorträge | Software | Person | Impressum, Datenschutzerklärung | Blog
Introduction
Historical Perspective
The history of computer graphics is short:
-
1963 Ivan Sutherland: Sketchpad (light pen etc.)
-
1975 Ed Catmull: z-Buffer (described one year earlier by Wolfgang Straßer
in his PhD thesis)
-
1980 Turner Whitted: Ray Tracing (described 1968 by Arthur Appel)
-
1986 Pixar: "Luxo, Jr."
-
1987...1989 Pixar: RenderMan and RenderMan Interface
Status quo of computer graphics:
-
We can generate images and motion pictures that look as if they had been
shot from reality. However, the rendering and even more the construction
of geometry, textures, and motions is very time-demanding. Use of this
technology for image or video production requires in-depth artistic and
technical skills.
-
We can generate interactive 3D worlds on a the screen of an inexpensive
computer. However, they mostly do not stand a close comparision to photographs.
Authoring is complex and typically requires highly developed programming
skills.
-
We can generate immersive 3D worlds using head-mounted displays, CAVEs,
etc. However, these types of devices still are too expensive to be used
much.
-
We are not able to generate an experience that comes close to a dream.
Trends:
-
Graphics hardware: closer to realism in real time
-
Motion synthesis, NPCs, etc.
On-going Discussions:
-
Alternatives to z-Buffer rendering for real-time applications: point-based
rendering, image-based rendering, raytracing
No Trends:
-
Affordable 3D scanning
-
Affordable motion capturing
-
VR for everybody
-
3D modeling, rendering, and a animation software without the steep learning
curve (despite some research in that direction)
Subdomains of Computer Graphics
(This is my very personal view. JL)
-
Modeling
-
Geometry
-
Polyhedra
-
Smooth surfaces (splines, subdivision surfaces, ...)
-
Volumes
-
...
-
"Look" of Surfaces
-
Reflection models
-
Textures
-
Measured surface properties
-
...
-
3D Reconstruction
-
Structured Lighting
-
Passive Vision
-
...
-
...
-
Animation
-
Physics
-
Keyframing
-
Deformation
-
Kinematics
-
Motion capturing and editing
-
Autonomous actors (cf. NPC = non-player characters): Game-AI
-
...
-
Rendering
-
Photorealism
-
Shadows
-
Indirect diffuse lighting
-
Caustics
-
...
-
Optimizations
-
Non-photorealistic rendering (NPR)
-
Point-Based Rendering
-
Image-Based Rendering
-
...
-
Everything else
-
Visualization, Data Mining
-
User interfaces
-
Image manipulation
-
Fonts, 2D, printing
-
Compositing, matting
-
Color management
-
Input and output devices
-
...
Much research focuses on these areas:
-
Rendering speed (offline/interactive/real time) vs. rendering quality
-
Minimize user effort in animation
-
Minimize user effort in modeling
Related Fields of Science
(in alphabetic order, incomplete)
-
Anatomy: body shape, kinematics, and dynamics
-
Artificial Intelligence: non-player characters
-
Audio Processing: sound generation and transmission in 3D worlds; 3D audio
-
Computer Engineering: graphics chips and subsystems, input and output devices
-
Computer Vision: 3D reconstruction
-
Game Research: game mechanics
-
Human-Computer Interface: WIMP, post-WIMP
-
Image Processing: filtering, antialiasing, HDR tone mapping, lossy compression
-
Mathematics: linear algebra, manifolds, partial differential equations,
optimization, etc.
-
Networking: multi-user games, distant rendering
-
Perception: lossy compression, color theory
-
Physics: simulations of rigid bodies, soft body, fluids; optical phenomena
-
Programming Languages: shading languages
-
Soft Computing: content-based retrieval
First Glimpse at Current Computer Graphics Research
Conferences
Nowadays, most important scientific works in computer graphics are published
on conferences for speed: Publications in journals may take longer than
a year from submission to print. A workaround that is used in computer
graphics: Some conference proceedings appear as journals.
Here are some conferences, sorted according to importance:
-
SIGGRAPH (conference of the ACM special interest group on computer graphics),
Eurographics (the European counterpiece of SIGGRAPH)
-
Symposium on Computer Animation, Symposium on Rendering, Symposium on Graphics
Hardware, Computer Graphics International, ...
-
Somewhat more local conferences such as SIBGRAPI, WSCG, SCCG, AFRIGRAPH,
Pacific Graphics, ...
Most of the conferences in computer graphics accept several kinds of contributions
such as full papers, short papers (SIGGRAPH: sketches & applications),
posters, educational papers, etc. Typical requirements in computer graphics:
-
No incremental, but only novel work
-
Proof-on-concept implementation with benchmark results etc.
-
Complete paper included already with the submission (The review will be
done based on that paper; in many other fields of science an abstract is
sufficient.)
Papers
-
Structure of a typical computer graphics paper. Examples.
-
Nowadays, most computer graphics papers are accompanied by a video. Examples.
-
What is possible in small-scale research? Examples from my work.
Remainder of this course
-
Topics
-
Ideas for software development coursework (please send me
an e-mail to receive the list)
-
Usable implementations of current research work
-
Proof-of-concept implementations of novel ideas
-
Lecture notes
-
Lab