Delta E formula match prediction

1 . A computer implemented method of determining a color formula for a target color, comprising:

generating a plurality of candidate color formulas to reproduce the target color by a color formulation engine;

for each candidate formula, determining a predicted delta E indicating a difference from a predicted color for the candidate formula from the target color;

for each candidate formula, the color formulation engine generating a confidence value in the predicted delta E by summing weighted figures of merit for each colorant in the candidate formula, wherein the weighting of the figures of merit represents each colorant's proportion in the candidate formula; and

ranking candidate formulas based on the predicted delta E and confidence value;

wherein the figure of merit for each colorant comprises an average standard deviation (PSD) of the colorant's predicted spectral response relative to spectral measurements of a plurality of samples of the colorant, where the samples are prepared at different colorant concentrations.

2 . The method of claim 1 , wherein the pSD for each colorant is obtained by:

obtaining spectral measurements for a plurality of samples made with the colorant at different colorant concentrations;

generating a plurality of formulas by generating at least one formulation to match the spectral measurement of each sample;

generating a predicted spectral response for each formula; and

determining an average standard deviation between the predicted spectral responses of the formulas and the spectral measurements of the samples.

3 . The method of claim 2 , wherein the color formulation engine generates the confidence value by:

determining a percentage of each colorant in the candidate formula;

multiplying each colorant's pSD by its percentage in the candidate formula to obtain a weighted pSD for each colorant;

summing the weighted pSDs; and

expressing the sum of the weighted pSDs as a percentage confidence.

4 . The method of claim 1 , wherein the target color is defined in terms of a multi-dimensional color space.

5 . The method of claim 1 , wherein the figure of merit further comprises an average delta E (pΔE) of a colorant's predicted color relative to spectral measurements of the colorant's samples.

6 . The method of claim 1 , further comprising indicating a best candidate color formula.

7 . The method of claim 1 , wherein the predicted delta E and confidence value comprises a plurality of predicted delta E and confidence values determined at a plurality of illumination or viewing angles.

8 . A system for determining a color formula to match a target color, comprising a computing device having non-volatile instructions that, when executed by a processor, cause the computing device to:

generate a plurality of candidate color formulas to reproduce the target color;

for each candidate formula, determine a predicted delta E indicating a difference from a predicted color for the candidate formula from the target color;

for each candidate formula, generate a confidence value in the predicted delta E by summing weighted figures of merit for each colorant in the candidate formula, wherein the weighting of the figures of merit represents each colorant's proportion in the candidate formula; and

indicate a formula to use based on the predicted delta E and confidence value;

wherein the figure of merit for each colorant comprises an average standard deviation (PSD) of the colorant's predicted spectral response relative to spectral measurements of a plurality of samples of the colorant, where the samples are prepared at different colorant concentrations.

9 . The system of claim 8 , wherein the pSD for each colorant is obtained by:

obtaining spectral measurements for a plurality of samples made with the colorant at different colorant concentrations;

generating a plurality of formulas by generating at least one formula to match the spectral measurement of each sample;

generating a predicted spectral response for each formula; and

determining an average standard deviation between the predicted spectral responses of the formulas and the spectral measurements of the samples.

10 . The system of claim 9 , wherein the confidence value is generated by:

determining a percentage of each colorant in the candidate formula;

multiplying each colorant's pSD by its percentage in the candidate formula to obtain a weighted pSD for each colorant;

summing the weighted pSDs; and

expressing the sum of the weighted pSDs as a percentage confidence.

11 . The system of any of claim 8 , wherein the target color is defined in terms of a multi-dimensional color space.

12 . The system of claim 8 , wherein the figure of merit further comprises an average delta E (pΔE) of a colorant's predicted color relative to spectral measurements of the colorant's samples.

13 . The system of claim 8 , wherein the wherein the predicted delta E and confidence value comprises a plurality of predicted delta E and confidence values determined at a plurality of illumination or viewing angles.

14 . A computer implemented method of determining a color formula for a target color, comprising:

generating a plurality of candidate color formulas to reproduce the target color by a color formulation engine;

for each candidate formula, determining a predicted delta E indicating a difference from a predicted color for the candidate formula from the target color;

for each candidate formula, the color formulation engine generating a confidence value in the predicted delta E by summing weighted figures of merit for each colorant in the candidate formula, wherein the weighting of the figures of merit represents each colorant's proportion in the candidate formula as generated by the color formulation engine; and

ranking candidate formulas based on the predicted delta E and confidence value;

wherein the figure of merit comprises an average delta E (pΔE) of a colorant's predicted color relative to spectral measurements of the colorant's samples.