Color appearance reproduction: visual data and predictive modeling

Abstract

Corresponding-colors data for complex images were obtained using an adjustment method. Experiments were carried out for printed images viewed at a high luminance in a light surround under CIE illuminant D50 simulators that were matched by CRT-displayed images using both CIE illuminant D65 and D50 white points at a lower luminance level in a dark surround. The results show expected chromatic adaptation effects, but indicate little or no surround and luminance level effects. The data were then used to evaluate various color-appearance models and chromatic adaptation transforms. It was shown that a simple linear transformation of tristimulus values between viewing conditions provided a good description of the data. The recently proposed CIECAM97s model performed among the best models when it was adjusted to account for the lack of a surround effect in these data.