Evaluating Textile Fabric Color Variation by Scanner

Abstract

The purpose of this publication is to describe the possibility of using a scanner for the evaluation of color variation or color difference of textile fabrics. Initially, the color specification of colored fabrics were measured by spectrophotometers and the actual color differences between colored fabrics were calculated by the δE*ab color difference formula. Then, a scanner was used to take images of the textile fabrics. The obtained images were filtered for noise removal. The RGB values of the obtained images were used for the evaluation of textile fabrics color variation. Several methods were used to evaluate color difference by a scanner. The best prediction was obtained by the neural networks methodwith 1.014 δE*ab. Using this method, the accuracy of prediction for training sets was better than that for the testing sets. The performance of each method depends on the color difference values. so that, for low and high color difference, the best prediction was obtained by neural network, and for median color difference, the best prediction was obtained by the multi linear regression method. The obtained results show that the color variation of textile fabrics can be estimated by using scanner RGB variation.