Design And Implementation Of Customized Textiles Platform Based On Cloud Services

The Internet has already penetrated into every corner of our lives, and textiles which are important daily necessities, are also turned to online trading. At the same time, the designing and producing of textiles are more effective since the textile CAD software is used widely in textile industry. However, most of current textile CAD softwares not only have high use cost but also are difficult to maintain. The cost of textiles is difficult to reduce for these reasons, and it is also hard to satisfy the requirement of customized textiles. In this thesis, the platform for customized textiles is presented.The platform is a vertical e-commerce platform for textile industry, which is implemented with Web technologies. And on the e-commerce platform, the information of textile industry is integrated, the textiles are customizable, and all the textiles can be traded online. It is obvious that the cost of textiles is lower since the sale links are less. Also, textile industry can get scale effect by integrating all the advantageous resources. Besides, the interaction experience is promoted by providing textile simulation and scene simulation for users, and the problem of high use cost and maintaining difficulties is solved by encapsulating the traditional textile CAD technologies into plugins of cloud services. The customized textiles platform is easy to extend if cloud services are applied in this way.The weaved photograph CAD technology is relatively difficult in all textile CAD technologies, while textile CAD technologies are the underlying technologies of the customized textiles platform. The algorithm of yarn selection and simulation on woven photograph is studied in this thesis.The algorithm of yarn selection is aimed to select a group of yarns from the yarn library, and make sure that this group of yarns can produce textile which has the minimum error compared with the original design. The simulated annealing genetic algorithm is adopted in this thesis, and the result shows that the algorithm can not only solve the problem of converging to local optimal solution but also improve the speed of converging. What’s more, the computational complexity of color separation is further reduced by eliminating the same colors and clustering the similar colors, and the clustering algorithm is used to compress the feasible solution area of yarn selection. In this way, the efficiency of yarn selection algorithm is improved. On the other hand, the algorithm of simulation is aimed to generate the preview image of textile by simulating the production of textile, so that users can get a perception of textile before producing, and designers can design better with the preview image. The thesis has realized two kinds of simulation:colorful simulation is generated by combining all the colorful slices, and the tissue simulation is generated by combining all the tissue slices, which are mapped from the true color slices with the mapping algorithm proposed in this thesis.