[Please remove <h1>]
Objectives
In computer graphics, 3D models are rendered through an ideal camera to a 2D image. The virtual camera is a powerful tool, controlling not only perspective and the optical path, but colour mapping as well. This course examines the notion of "colour" at different stages of the computer graphics pipeline, and describes techniques for modelling and managing colour through the virtual camera analogy.
Topics include illumination, the virtual camera, colour spaces, gamut mapping
and colour management. Implementation of the colour reproduction and gamut-mapping
algorithms demands a number of numerical and statistical methods, including
multidimensional interpolation and approximation, regression, computational
geometry, tensor-product and simplex splines, principal components analysis,
and linear and non-linear optimization. Programming assignments in colour reproduction
will be required, as well as a final project on a relevant research topic. Students
taking this course should have a strong background in computer graphics and
numerical methods.
References
Illumination and Color in Computer Generated Imagery, Roy Hall, Springer-Verlag,
1989. Visual Computing: the Integration of Computer Graphics, Visual perception,
and Imaging, Markus Gross, Springer-Verlag, 1994. Computer Graphics:
Principles and Practice, James D. Foley, Andries van Dam, Steven K. Feiner,
John F. Hughes, Addison-Wesley, 1990 (Ch. 4, Graphics Hardware and ch. 13, Achromatic
and Coloured Light). Principles of Digital Image Synthesis, Andrew S. Glassner,
Morgan Kaufmann, 1995 (Ch. 1, The Human Visual System and Color, and Ch. 11,
Light). Color Technology for Electronic Imaging Devices, Henry R. Kang,
SPIE Optical Engineering Press, 1997. Light and Color in Nature and Art,
Samuel J. Williamson and Herman Z.Cummins, Wiley, John and Sons Inc., 1990.
Schedule
3 hours of lectures per week. Normally offered every other year.
Outline
Illumination (6 hrs)
Basics of light and materials. Data structures for spectral power distributions (SPDs), surface geometry, reflectance, and transmittance.
The Virtual Camera (6 hrs)
Rendering and the virtual camera. Algorithms for manipulating the optical path, trichromatic and spectral colour mapping, comparisons with digital camera technology.
Perceptual Response and Colour Spaces (6 hrs)
Mathematical models of the human visual system and trichromatic colour spaces, the CIE standard observer. Algorithms for colour mapping in high-dimensional spaces, linear reflectance models. Principal components analysis (PCA) and singular value decomposition (SVD).
Image Display (6 hrs)
Data structures for printer and monitor gamuts. Statistical methods for fitting colour data, multidimensional interpolation and approximation, regression.
Gamut Mapping (6 hrs)
Gamut-mapping algorithms: black- and white-point mapping, neutral axis alignment, advanced methods. Linear and non-linear optimization for out-of-gamut colour projection. Computational geometry of device gamuts. Tensor-product and simplex spline methods for modelling and gamut mapping.
Colour Management and Operating Systems (6 hrs)
Colour management algorithms in various operating systems. International Color Consortium (ICC) profiles, ColorSync, Kodak Color Management System (KCMS), ANSI IT8 targets.