Page 94 - Computational Colour Science Using MATLAB

6

Chromatic-adaptation Transforms

and Colour Appearance

6.1 Introduction

The distinction between colour specification and colour appearance was touched
upon in the review of the CIE system presented in Chapter 1, Section 1.3.
Whereas the CIE system of colorimetry (based upon XYZ tristimulus values) is
clearly a system for colour specification, some advanced colour specification
models such as CIELAB could arguably be described as models of colour
appearance. The polar coordinates of CIELAB allow the description of a colour
stimulus in terms of three terms, lightness, chroma and hue, and these correlate
quite well with the perceptual attributes of brightness, colourfulness and hue.
Furthermore, the normalization procedures inherent in the transform from XYZ
to L*a*b* result in a* and b* values close to zero for a perfect reflecting diffuser
(or any surface whose spectral reflectance does not vary with wavelength)
irrespective of the illuminant. This is consistent with the fact that surfaces in
general tend to retain their colour appearance when viewed under a wide range
of light sources and contrasts with the properties of the XYZ system. CIELAB is
a relatively poor colour-appearance model, however, and this chapter describes
several advanced colour-appearance models (CAMs). The human visual system
has a remarkable ability to maintain the colour appearance of an object despite
quite large changes in the quality and intensity of the illumination. A white piece
of paper tends to look white whether it is viewed by daylight, Tungsten light or
candle light. It is generally considered that the human visual system achieves
colour constancy by some process that allows it to discount the effect of the
illumination. The term chromatic adaptation is often used to describe this
process and the chromatic adaptation is said to be complete if the effect of the
illumination is completely discounted. Most CAMs therefore include a
chromatic-adaptation transform (CAT). A CAT is a method for computing

Computational Colour Science Using MATLAB. By Stephen Westland and Caterina Ripamonti.
& 2004 John Wiley & Sons, Ltd: ISBN 0 470 84562 7