Wine Color Characterization and Classification for Nuances Reproduction

In this paper, we propose a simple method for wine color characterization, classification and reproduction. The aim is to represent the colors of wines with limited number of hues that we call nuances. Burgundy wines (France) constitute the wine samples in this study but the method remains general. The method consists of four steps: spectral transmittance measures of a large number of wine samples. Then standard and gamma corrected colors are reconstructed from spectral data. Afterwards, a ΔE-based classification is performed in the CIELAB, which provides good visual uniformity and thus offers the best discrimination between the different samples. The last step is a spectral-based color reproduction using synthetic liquids. The obtained results are encouraging in that they permit an accurate characterization and reproduction of wine color.