Study on color representation model and computer simulation of colored spun fabrics based on image translation model

Abstract

To improve the design and development efficiency of colored spun products, this paper puts forward a computer simulation scheme based on an image translation model, focusing on the ‘fancy’ coloration characteristics of colored spun fabrics. First, the Pix2PixHD image translation model was used to train and extract the fabric color texture information of colored spun fabrics, so as to construct a color texture characterization model; meanwhile, the texture fusion and color filling algorithm was constructed for the blending process of two-component dyed fibers, so as to realize the computer simulation of the two-component colored spun fabric ‘fancy’ color effect. On this basis, a texture and coloration characteristic evaluation index for colored spun fabric was established, which was applied to the quantitative analysis of computer simulation results for the first time. Meanwhile, the difference significance test was performed on the quantitative analysis results. It turned out that the computer simulation scheme established in this paper can effectively simulate the various ‘fancy’ color effects. Compared with the existing algorithms, it not only has obvious advantages in visual subjective feeling, but also has a notable improvement in objective and quantitative analysis data. The texture similarity and color similarity are increased by 5.78% and 19.17%, respectively, the peak signal-to-noise ratio and structural similarity index measure are increased by 15.72 db and 0.168, respectively. Therefore, the computer simulation scheme has ideal realistic and generalization characteristics.