C0S 597d - Advanced Topics in CS

Sensing for Graphics

Fall, 2001


Course home      |      Outline      |      Paper signup   1   2   3

Summary

The course will be organized around four major data types that may be measured from the real world:

For each, we will discuss:

The following detailed schedule should be regarded as preliminary and likely to change.


Course Outline

Monday,
September 17
  • Introduction: real-world data in computer graphics
  • Plenoptic function: definition, taxonomy
  • Sources of error in acquired images
Wednesday,
September 19
  • Image formation, lenses, focus, geometric distortion
  • Image sensors (CCD, CMOS), dynamic range
  • Color and chromatic aberration
  • Slides available as PPT or as images.
Monday,
September 24
Wednesday,
September 26
Monday,
October 1
  • Intro to image-based modeling and rendering
  • Lightfields and lumigraphs
  • Background reading:
  • Slides available as PPT or as images.
Wednesday,
October 3
Monday,
October 8
  • Taxonomy of reflectance and scattering measurements
  • The BRDF: definition, models, representation
  • Other data types: BTF, BSSRDF
Wednesday,
October 10
  • Guest lecture by Daniel Aliaga:
    "Reconstructing 3D Interactive Walkthroughs from Omnidirectional Video"
Monday,
October 15
Wednesday,
October 17
Monday,
October 22
Wednesday,
October 24
Monday,
October 29
  • Fall break - no class
Wednesday,
October 31
  • Fall break - no class
Monday,
November 5
Wednesday,
November 7
  • Taxonomy of methods for acquiring 3D shape
  • Brief overview of passive methods
  • Active range scanning: touch probes, defocus, time-of-flight, photometric stereo, structured-light triangulation
  • Slides available here.
Monday,
November 12
  • Guest lecture by Claudio Silva on surface reconstruction
Wednesday,
November 14
  • Practical issues in the design of 3D laser scanners
  • Scan data postprocessing pipeline: scan alignment, global registration, merging, surface reconstruction, holefilling, etc.
Monday,
November 19
Wednesday,
November 21
  • Thanksgiving - no class
Monday,
November 26
Wednesday,
November 28
Monday,
December 3
  • Motion capture technologies: mechanical, inertial, ultrasonic, magnetic, optical (marker-based and markerless)
Wednesday,
December 5
Monday,
December 10
Wednesday,
December 12
  • Presentations of student projects and course wrapup


Last update 12:05:46 29-Dec-2010
smr@cs.princeton.edu