Computer Colour Matching and Recipe Prediction
topic
Computer colour matching software applies Kubelka-Munk two-flux optical theory to model reflectance of dye mixtures from single-dye calibration data, using non-linear optimisation to calculate dye concentrations producing minimum colour difference from target reflectance; multi-solution ranking by metamerism, cost, and reproducibility criteria selects the commercially optimal recipe from the theoretically equivalent solutions.
Role
Transforms textile dyehouse colour recipe development from empirical trial-and-error iteration to model-based computational prediction that provides starting recipes within commercial tolerance on the first or second laboratory trial, reducing dyehouse recipe development time and raw material consumption while improving first-time-right colour matching rates across diverse substrate and dye class combinations.