A pattern intelligent typesetting method for textile printing and dyeing products

Abstract

The invention discloses an intelligent type setting method for flower patterns of a textile printing and dyeing product. The intelligent type setting method is an automatic flower pattern type setting and generating method based on a neural network depth learning technology. By learning on a large number of existing flower pattern design products, flower pattern materials and a geometric incidence relation between the materials are automatically extracted; scoring and grading on material combinations is performed by adopting a probability analysis model; and next, other materials are selected intelligently according to seed materials selected by a user, and a plurality of flower pattern material type setting schemes are generated. Therefore, the design difficulty and design period of flower pattern type setting are greatly lowered, and the attractive appearance and types of the flower patterns are improved.

Description

technical field [0001] The present invention relates to the fields of digital printing and dyeing, neural network, deep learning, and image recognition, in particular to an intelligent typesetting method for pattern of textile printing and dyeing products, which is based on neural network pattern learning and prediction. Background technique [0002] Digital printing and dyeing is the development direction of the textile industry in the future. Its biggest feature is that it can support the production of highly customized printing and dyeing products. The unit price of fabrics of 1 meter and 1000 meters is the same; while the traditional printing and dyeing needs to print hundreds of sheets at a time. The cost can only be recovered if it is more than one meter, so it is impossible to support customized products. Digital printing and dyeing in developed countries has accounted for 48% of the market, while my country only accounts for 2%, so there is huge room for development....

AI Technical Summary

Problems solved by technology

[0003] However, the pattern design of printing and dyeing products has a high threshold, high difficulty, time-consuming and laborious, and usually takes a few days to complete by professional designers, which greatly hinders users' interest in customizing their own printing and dyeing products

However, automatic pattern design is extremely difficult, and there are no successful cases to follow

[0004] The aesthetics of pattern selection and design typesetting are very abstract subjective descriptions. Traditional image recognition algorithms, such as algorithms based on image sift features and matching algorithms based on manifolds, cannot accurately describe the quality of typesetting.

Using learning algorithms such as decision trees and Bayesian classification, it is also impossible to evaluate the quality of patterns

Images