Ten unsolved problems in computer graphics

1
Ten unsolved problems in computer graphics

• Ivan Sutherland
• Presented by Lee Jaehyung
• CS Dept. KAIST
• gloine_at_bawi.org

2
Introduction

• First suggested by Ivan Sutherland in 1965.
• A course at Univ. of Michigan in June 1965.
• I.E. Sutherland, Ten Unsolved Problems in
  Computer Graphics, Datamation, Vol. 12, No. 5,
  May 1966, pp. 22-27.
• Became the focus of future developments in 70s
  and 80s.

3
Problem List

• Hardware characteristics and cost
• Problems of technique (e.g. rubber banding)
• Coupling problems (display-to-simulation)
• Describing motion
• Digital halftoning
• Structure of drawings (making the structure
  explicit)
• Hidden line removal
• Program instrumentation (and visualization)
• Logical arrangement (e.g. hierarchical modeling)
• Working with abstractions (e.g. scientific
  visualization)

4
Hardware characteristics and cost

• Cheap machines with basic capability
• The lack of a complete graphical console at
  reasonable cost
• Invention of raster displays, frame buffer, and
  VLSI technology.
• Today Cheap PCs with 3D accelerators!

5
Problems of technique

• Making graphics friendly
• For example, rubber banding technique used in
  some programs such as winamp.

• Today Techniques are invented in personal
  computers

6
Coupling problems

• Flight simulator
• Visual data displayed by the simulator must be
  coupled with the behavior of a pilot .
• It is difficult to provide exact vision and
  motion cues on the ground.
• In-flight simulation
• original dynamic response additional control
  system from the data gathered by real aircraft
• Today Detecting eye gaze direction, etc.

7
Describing motion

• Researchers had still to develop languages that
  included motion capability in their semantic
  structure.
• The motion language by Kenneth Knowlton at Bell
  Labs. (BEFLIX and EXPLOR)
• Used to produce animated films.
• Drew upon laws of physics that are still used in
  producing animated computer graphics.
• Today script languages in 3DS and Maya

8
Digital halftoning

• Definition the process of using dots of varying
  size or arrangements (rather than varying
  intensity) to form varying degrees of a color.
• patterning, dithering, error diffusion
• Today We can find recent papers on this topic!

9
Structure of drawings

• Extending Lawrence Roberts's work from three
  dimensional solid objects to arbitrary curved
  surfaces
• Roberts, L. G., Machine Perception of
  Three-Dimensional Solids, Tippett, J. T. (ed.),
  Optical and Electro-Optical Information
  Processing, MIT Press, Ch. 9, Cambridge,
  Massachusetts, p159-197, 1965.
• process a photograph into a line drawing,
  transform the line drawing into a
  three-dimensional representation, and ,finally,
  display the three-dimensional structure with all
  the hidden lines removed, from any point of view.
• Today It is a pattern recognition task!

10
Hidden line removal

• For wire-framed images.
• hidden surface removal problem, with raster
  displays.
• In 1974, Sutherland, Robert Sproull and Robert
  Schumaker, published a paper titled A
  Characterization of Ten Hidden Surface
  Algorithms.
• to compare the available hidden-surface
  algorithms and synthesize a greater understanding
  of the problem.
• Underlying principles sorting and coherence
• In 1974, Edwin Catmull, "bivariate patches
• associated with Coons's earlier work on surface
  patches.
• z-buffer system for hidden surfaces.
• Today can we have more efficient algorithms?

11
Program instrumentation

• To visualize execution of concurrent programs.
• DARPA'S Defense Sciences Office in 1980s
• Computer systems had become so large and
  complicate that no single person could know all
  about any one system.
• A need for efficient and effective ways each
  system had to include ways for users to access,
  make changes to, and update the information.
  (documentation was ineffective!)
• An interactive way to visualize a system was
  needed.
• Today PV in IBM research, Numega Devpartner,
  etc.

12
Logical arrangement

• hierarchical modeling in todays term.
• By Sutherland Presenting information in a
  sensible or artistic or logical arrangement
• Today Every 3D graphics tool supports hierarchy.

13
Working with abstractions

• He called for what later would be known as
  scientific visualization (SciVis).
• To understand the data.

• Today Many SciVis tools such as Data
  Visualizer, Iris Explorer, Data eXplorer, etc.
• Free tool Mayavi

14
Conclusion

• They were guidelines of research in 1970s and
  1980s.
• Some problems are still not solved enough.
• Today there are many other problems to solve.
  (e.g. antialising)
• J. F. Blinn, Ten more unsolved problems in
  computer graphics, IEEE Computer Graphics and
  Applications, Vol. 18,  Issue 5, Sep. 1998, pp.
  86-89.

15
Reference

• http//graphics.stanford.edu/courses/cs348b-98/ope
  n_problems.html
• http//www.stanford.edu/group/mmdd/SiliconValley/N
  orberg/TransformCompTech/Norberg_Chapter3.html
• J. F. Blinn, Ten more unsolved problems in
  computer graphics, IEEE Computer Graphics and
  Applications, Vol. 18,  Issue 5, Sep. 1998, pp.
  86-89