Visual pattern auto-generation for yarn-dyed plaid fabric based on modified genetic operators

Abstract

With the aggravated homogenization of the commodity in the increasingly competitive marketplace, the industries have been converted to design and manufacture small batch productions with a wide variety. To reduce the designer’s burden and shorten the design cycle of the product, an auto-generation method for the pattern of yarn-dyed fabric was proposed based on the modified genetic operators. Initially, the design elements to form the pattern were extracted and encoded as a chromosome by natural number encoding. Then, the fabric weave model was built using the matrix of 0 and 1. Based on the color scheme and yarn arrangement, the pattern of yarn-dyed fabric can be generated automatically. The elements can be exchanged to generate new patterns by the genetic operators. Finally, the obtained pattern was imported into the three-dimensional virtual shirt model. Experiments show that the proposed method can help the designer obtain different patterns quickly and effectively, and different design elements can be controlled according to the requirements of the designer. Based on the annual fashion trend, fashionable colors can be involved to realize fast reaction in product design.