Prediction of Yarn Cross-Sectional Color from Longitudinal Color by Neural Network

Abstract

This paper presents a study of color behavior and reflectance factors of fibers in their crosssectional and longitudinal directions. In the practical field, it was demonstrated that the reflectance factors and the lightness (L*) of fibers are different in their cross-sectional and longitudinal directions and the mentioned values are higher for the lateral surface. The measurement of reflectance factors of fibers' cross-sections showed that lightness and reflectance factors increase as fibers’ density increase in the measurement cell. A relationship was established between colors of fibers in two directions using the ratio of K/S values of yarns in cross-sectional and the longitudinal directions at different wavelengths. In the second part of this work a neural network was applied to relate the color of fibers in the mentioned directions. Results showed an excellent prediction with the later technique.