Premium
Vector‐based modelling of colour difference: a pilot study of the DE 2000 colour difference model
Author(s) -
Oulton David P.,
Westland Stephen
Publication year - 2017
Publication title -
coloration technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.297
H-Index - 49
eISSN - 1478-4408
pISSN - 1472-3581
DOI - 10.1111/cote.12242
Subject(s) - mathematics , hue , color difference , colour difference , significant difference , ellipse , equivalence (formal languages) , dimension (graph theory) , lightness , constant (computer programming) , mathematical analysis , rgb color model , combinatorics , geometry , statistics , artificial intelligence , optics , pure mathematics , computer science , physics , enhanced data rates for gsm evolution , programming language
A novel approach to colour difference modelling is presented whereby for any given CMC (1:1) or CIE DE 2000 ∆ E , ∆ C , ∆ H , and ∆ L colour difference, the equivalent CIE XYZ , L*a*b* , and L*C*h coordinate changes are derived by optimising the input RGB stimuli from which they are all calculated. Single‐dimension L or C or H difference loci expressed in DE 2000 difference units are thus generated, and the additive equivalence of tristimulus values is likewise projected forward onto each locus and also onto a set of CIE DE 2000 three‐unit ellipse boundaries. Using the datasets thus generated, it is then shown firstly that the derived ellipses have well‐defined semi‐axes, which explain the detailed orientation of the MacAdam ellipses in x , y , Y space. Unit CIE DE 2000 difference is confirmed as a successful quantifying constant of visual difference over a wide range of chroma, hue, and lightness differences. As a constant, CIE DE 2000 unit difference is shown to have a significantly variable value at high and low chroma: evidence is established for systematic changes in both chroma and hue difference sensitivity. A hitherto unresolved non‐linearity is revealed in the C* dimension of L*C*h space that is not replicated in the CIE DE 2000 model. The derived difference loci appear to specify physically reproducible experimental stimuli that could be used in the estimation of visual difference magnitude. Overall, the data derived by the new approach and presented in this paper increase the probability that a true vector model of the visual difference response may eventually be derived.