Research on color transfer law of color spinning process based on multi-dimensional feature analysis

Taking the color characteristics among dyed polyester fibers, yarns and woven fabrics in color spinning processing as the research object, this paper mainly discusses the influence of process parameters and tissue structure on their color transfer during the yarn formation and weaving processes. Aiming at the limitation of finite point measurement, a multi-dimensional feature analysis was conducted by the spectral pan-similarity measure model and chromaticity index. Combined with the color difference between samples and their own color fluctuation, the relative difference criterion is introduced to analyze the color transfer law between colored spun samples. The results show that the spectral difference between polyester strips and color spun yarns are all greater than those between colored spun yarns, indicating that the morphological changes between polyester strips and yarns have a more pronounced effect on color transfer than the twist factor variations between colored spun yarns. In addition, the spectral difference between colored spun yarn and its satin fabric is the smallest, followed by twill and plain is the largest. Furthermore, the spectral-based feature analysis presents the same law as the color difference feature analysis, but the color difference can not distinguish the color presentation difference between some samples. The relative discrimination criterion can significantly enhance the validity and accuracy of the evaluation index.