Study On Automatic Generation Of Female Fitting Jeans Pattern

It is that rapid development of the digital production process of clothe,and the fashion enterprise pattern has also been developed mainly by the garment CAD technology,and with the development of the customization industry,the personalize customization and the mass customization of the consumer group are increased year by year.Jeans,as a basic popular single item,occupies a dominant position in the sales of single item of clothing,among which full-body jeans are also an essential single item in People's Daily life.In the market,elastic denim fabrics are mostly used to meet the needs of various types of jeans for women's body-fitting jeans,but it is really difficult to meet the characteristics and needs of various types of body-fitting jeans for non-elastic body-fitting jeans,so most of them are produced in a customized way.In this form,it has to be adjusted for the size of each individual to make the time and cost of the production,against the production and development of the enterprise.In view of this situation,this paper proposes a GA-BP neural network is adopted to establish from the human body size to fit jeans size sample detail size of reasoning model,the model of jeans in parameterization processing,use of Auto CAD secondary development environment to write automatic plate making process,thus completed the ms for mass customization model automatic generation system combined size of jeans.First of all,we visited the garment mass customization enterprises to understand the market situation,conducted market research on full-body jeans,and determined that the main consumer groups were young women.Refer to the relevant literature of jeans sample research,sample automatic generation technology and neural network for analysis and arrangement.Through the joint research with the expert of clothing pattern making and experienced pattern making master,the basic pattern and pattern making rules of women's fit jeans are determined.TC2 was used to measure 198 sets of human body data,sort out and analyze the human body data,and finally,the193 group of data samples were reserved and extract the size of each part of the human body necessary to draw the jeans template.Then,the professional plate maker will draw the jeans sample according to 193 sets of human body data,measure the detail size,record and sort it out,and serve as the training data sample of the model of reasoning the sample size.Secondly by building detail model based on BP neural network's jeans size inference model,found that the iterative 147 times to reach the ideal error control,in order to improve the efficiency of inference model,this paper USES the genetic algorithm of BP neural network randomly generated initial weights and threshold optimization,build the GA-BP neural network as the model of reasoning model.Through network training and testing,a stable expected error was achieved after 73 iterations,and its adaptability reached 0.98937 with good results.Then,parametric design is carried out for the plate making process of fitted jeans.In the fitting of structural curve,the side seam and crotch curve of pants are fitted mainly by cubic spline curve and Bezier curve,and b-spline curve is used for adjustment to conform to the human body.In the secondary development of Auto CAD,the Auto LISP development environment of Auto CAD was selected to write the program of parametric automatic template drawing.Finally,the detailed size of the template derived from ga-bp was filled in the Auto CAD dialog box to automatically draw the corresponding jeans template.Finally,the practical verification of the above automatic generation system of women's body shape pattern was carried out,which was respectively verified from three aspects: technical theory,3D virtual fitting and real person garment fitting.The verification of technical theory is mainly based on the rationality and joinability of the template.3D virtual verification mainly USES 3D virtual fitting software to set the corresponding human body size to generate virtual models,sew and try on the samples.The fitting effect is mainly evaluated and analyzed from the effect diagram,perspective diagram and pressure diagram.The fitting of real clothes is mainly evaluated by experts and models themselves,which can be divided into fit evaluation and wear comfort evaluation,and then the comprehensive evaluation and analysis are carried out.The results show that the fit and comfort of the jeans sample automatically generated by the system meet the basic wear requirements,and the system has a high availability,can be used for mass customization of clothing system,can greatly improve the platemaking and production efficiency.