Abstract
This paper aims to solve the bottleneck problem of the K-M prediction algorithm in a multidimensional color-mixing system. First, according to the needs of full-color-gamut color matching, nine primary color fibers that can regulate hue, saturation, and lightness were preferentially selected, and based on the ternary double-coupling color-mixing mode, 18 ternary double-coupling grid color-mixing models were constructed; these were combined to form the cylindrical full-color-gamut grid color-mixing model that contains three planes with different lightness. Second, the K-M prediction algorithm was reconstructed to extend its sample space from a one-dimensional linear space to a three-dimensional space, and the color prediction of rotor-blended yarns was performed based on it. Finally, 273 uniformly distributed grid points in the full-color-gamut grid color-mixing model were selected as training and validation samples to prepare blended yarns and knitted fabrics, based on which the K-M color-prediction algorithm for the full-color-gamut grid color-mixing model was constructed and validated. The results showed that the mean value of the predicted color difference for the 18 validation samples was 0.691, and the mean value of the prediction error for the primary color fibers blending concentration was 1.73%. It is indicated that the constructed prediction algorithm realized the accurate prediction of color values and blending concentration of primary color fibers of full-color-gamut blended yarns.
Keywords
Get full access to this article
View all access options for this article.