Premium
The Application of Computer Color Matching Techniques to the Matching of Target Colors in a Food Substrate: A First Step in the Development of Foods with Customized Appearance
Author(s) -
Kim Sandra,
Golding Matt,
Archer Richard H.
Publication year - 2012
Publication title -
journal of food science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.772
H-Index - 150
eISSN - 1750-3841
pISSN - 0022-1147
DOI - 10.1111/j.1750-3841.2012.02744.x
Subject(s) - gamut , tile , color difference , matching (statistics) , mathematics , rgb color model , color model , colorimetry , computer science , substrate (aquarium) , color space , artificial intelligence , biological system , pattern recognition (psychology) , computer vision , materials science , statistics , image (mathematics) , geology , enhanced data rates for gsm evolution , composite material , biology , oceanography
A predictive color matching model based on the colorimetric technique was developed and used to calculate the concentrations of primary food dyes needed in a model food substrate to match a set of standard tile colors. This research is the first stage in the development of novel three‐dimensional (3D) foods in which color images or designs can be rapidly reproduced in 3D form. Absorption coefficients were derived for each dye, from a concentration series in the model substrate, a microwave‐baked cake. When used in a linear, additive blending model these coefficients were able to predict cake color from selected dye blends to within 3 ΔE* ab,10 color difference units, or within the limit of a visually acceptable match. Absorption coefficients were converted to pseudo X 10 , Y 10 , and Z 10 tri‐stimulus values (X 10 P , Y 10 P , Z 10 P ) for colorimetric matching. The Allen algorithm was used to calculate dye concentrations to match the X 10 P , Y 10 P , and Z 10 P values of each tile color. Several recipes for each color were computed with the tile specular component included or excluded, and tested in the cake. Some tile colors proved out‐of‐gamut, limited by legal dye concentrations; these were scaled to within legal range. Actual differences suggest reasonable visual matches could be achieved for within‐gamut tile colors. The Allen algorithm, with appropriate adjustments of concentration outputs, could provide a sufficiently rapid and accurate calculation tool for 3D color food printing. Practical Application: The predictive color matching approach shows potential for use in a novel embodiment of 3D food printing in which a color image or design could be rendered within a food matrix through the selective blending of primary dyes to reproduce each color element. The on‐demand nature of this food application requires rapid color outputs which could be provided by the color matching technique, currently used in nonfood industries, rather than by empirical food industry methods.