A spectral reflectance-based approach for formulating color mixing recipes for single- and multi-layered woven structures

The color of yarn dyed woven fabrics comes from a series of different colored yarns mixed in complex ways and various proportions. Thus, predicting the color mixing effect or formulating the recipe is a difficult task which should consider the interaction between colored yarns and structure variations. Color mixing recipes, which are also called color prediction models, for woven fabrics have generally been derived through the two-dimensional modeling of woven structures. In this study, the three-dimensional geometrical and colorimetrical modelings of single-, double-, and three-layered woven fabrics in a wide range of colors were conducted to obtain two types of optimal spectral reflectance-based color prediction models. Through model evaluations, the obtained three-dimensional color prediction models were proved to have much higher predictive accuracy, especially in terms of lightness and chroma predictions, than that of the two-dimensional color prediction models previously developed.