CIE
Division 8 - TC8-03: Survey of Gamut Mapping Papers
Braun & Fairchild (1998)
The aim of this paper is to investigate the effect of correcting hue non-uniformities
in CIELAB on the performance of gamut mapping algorithms implemented in
it. Two hue-linearized versions of the CIELAB color space were generated
to this end, whereby one was based on the Hung and Berns (1995) visual
data and one on the Ebner and Fairchild (1998a) data set (both data sets
consist of visually mapped hue data to planes of constant visual hue).
These modified versions of the CIELAB color space were psychophysically
tested for their hue-linearity characteristics against the CIELAB color
space. The results of these experiments show that, in the "blue" region
of CIELAB, the hue-corrected color spaces are more visually uniform and
perform better than CIELAB in gamut mapping situations with respect to
perceived hue. However, the CIELAB color space performed as good as or
better than either hue-corrected spaces outside of the blue region.
The gamut mapping algorithms used in this study were the following three:
1. Chroma clipping with lightness and hue preservation,
2. Minimum E*ab clipping with hue preservation, and
3. Centroid clipping towards L*a*b* = [50,0,0] with hue preservation.
However, as the study was aimed at investigating hue-linearity, no information
is provided about the relative performance of the GMAs used.
Last updated: 17 August 1999 by Jan
Morovic