Premium
Comparison of colorimetric and spectrophotometric algorithms for computer match prediction
Author(s) -
Sluban Boris
Publication year - 1993
Publication title -
color research and application
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.393
H-Index - 62
eISSN - 1520-6378
pISSN - 0361-2317
DOI - 10.1002/col.5080180204
Subject(s) - standard illuminant , matching (statistics) , curve fitting , mathematics , colorimetry , algorithm , computer science , least squares function approximation , artificial intelligence , statistics , computer vision , estimator
The article discusses two types of colour‐formulation strategies: the tristimulus colorimetric strategy (ΔX,ΔY,ΔZ) → (0,0,0) and the least‐squares curve fitting Σ j w 2 j [ΔR(λ j )] 2 → min. Then a new matching strategy (ΔE D65 ) 2 + (ΔE A ) 2 + (ΔE WWF ) 2 → min is stated, which tries to combine the advantages of the first two strategies: more “balanced” colour differences for several different illuminants (the case of curve fitting) and small colour differences for the illuminant involved in the matching process (the case of tristimulus matching). An algorithm following this new general strategy is constructed using the case of single‐constant Kubelka‐Munk theory as an example. Results of a few numerical experiments are included for illustration.