Don Norman

1
Don Normans Fall 2003 Seminar

• 8-Dimensional Digital Photography
• for Historical Preservation

Jack Tumblin jet_at_cs.northwestern.edu
2
Museum Challenge

• The Vast Lonely Warehouse, tarved
• Share, Preserve,
• and Educate
• (aquire, restore, explore)
• Cant we dobetter than just webphotos?

3
Better Other, Multiple Viewpoints?

• Paul Rademacher, Multiple Center of Projection
  Images SIGGRAPH1998

4
Better Variable Lighting?
Find just 6 coeffs/pixel
Get Interactivesynthetic re-lighting...
Malzbender,HPlabs Polynomial Texture Maps
SIGG2001
5
A Complete Optical Archive ?

• SIGGRAPH 1996 4D Light Fields or Lumigraph

Levoy et al.
Gortler et al.
Camera

• Angular
• Resolution
• at camera?
• at surface pt?

6
How An 8-D Appearance Field

• Coupling Between
• Orthographic camera, (x,y)positioned on sphere,
  (?, ?)
• and
• Orthographic projector, (x,y)positioned on
  sphere (?, ?).
• F( xc, yc, ?c, ?c, xp, yp, ?p, ?p, ?, t)

camera
?c
?c
7
How An 8-D Appearance Field

• Coupling Between
• Orthographic camera, (x,y)positioned on sphere,
  (?, ?)
• and
• Orthographic projector, (x,y)positioned on
  sphere (?, ?).
• F( xc, yc, ?c, ?c, xp, yp, ?p, ?p, ?, t)

camera
?c
?c
?p
?p
Projector (laser brick)
8
Conclusions

• Unified Ray Archive
• More photos? complete
• Preposterously huge, redundant
• Many new tricks needed

9
end
10
Seitz View Morphing SIGG96

• http//www.cs.washington.edu/homes/seitz/vmorph/v
  morph.htm
• Mix 2D and
• VERY SPARSE
• 3D data

11
Seitz View Morphing SIGG96

• http//www.cs.washington.edu/homes/seitz/vmorph/vm
  orph.htm

12
Seitz View Morphing SIGG96

• http//www.cs.washington.edu/homes/seitz/vmorph/vm
  orph.htm

13
Seitz View Morphing SIGG96

• http//www.cs.washington.edu/homes/seitz/vmorph/vm
  orph.htm

14
Seitz View Morphing SIGG96

• http//www.cs.washington.edu/homes/seitz/vmorph/vm
  orph.htm

15
Scene causes Light Field

• Light field holds all outgoing light rays

Shape, Position, Movement,
Emitted Light
Reflected, Scattered, Light
BRDF, Texture, Scattering
Cameras capture subset of these rays.
16
? Can we recover Shape ?

• Can you find ray intersections? Or ray depth?

17
? Can we recover Surface Material ?

• Can you find ray intersections? Or ray depth?

Ray colors might not match for non-diffuse
materials (BRDF)