Study On Automatic Generation System Of Casual Cheongsam Under The Background Of EMTM

With the improvement of consumers' requirements about fitness and comfort level of clothing structure,the emergence of high-quality and personalized clothing consumption have encouraged the development of industrialized customization.With the continuous improvement of modern society and economy,more and more customers prefer this way.Nowadays,since the traditional culture has been renewed attention,cheongsam is also regarded as one of the symbols of the Chinese nation.Not only the leaders' wife,female diplomats in foreign affairs activities wearing a cheongsam,but also those hostesses work in large international conference and sports event.Under the background of the Internet + Era,it meets the needs of the market and the tide of the times by combining the Chinese dress with digital.It uses the single form of customized production because of its high fitness characteristics.Because of its long production cycle,the process is complicated and not adapt to the modern way of life.So,is there any possibility to achieve mass customization,even the automatically generated customization pattern?This thesis introduces the background and research states of automatic generation system of casual cheongsam,at the meantime;it also analyzed pattern research,pattern automatic generation system and research status at home and abroad,which was the theoretical basis for the follow-up study.The second chapter introduces the main style of cheongsam in today's market and introduces the concept of life dress cheongsam.Through the method of literature review,we can determine the specific style of life dress and the role of the body according to the characteristics of the main influence.This study used Boss-21 three-dimensional body scanner to generate the pattern's key parts of data collection.And then analyzed and compared the advantages and disadvantages between the short-cut method,the prototype method,the proportional method and the three-dimensional cutting method.It selected the short-inch method for the plate-making by combining with the specific need,that is custom-made.It has laid the foundation for the follow-up model of the Chinese dress model according to determining the structure design steps of the Chinese dress.The third chapter established the pattern specification model and set up the mapping relationship between the actual size and garment size by BP neural network,selected the most reasonable neural network to design structure and parameters after the comparison and analysis.Then the structure of the neural network was trained to make the sum of squared error reach the standard range.The fourth chapter did a detailed analysis of the Chinese dress intelligent model's realization.Firstly,introduced the fitting theory of different function curves,analyzed and compared spline curves,Bezier curves and B-Spline,which influenced different parts of the spline curve fitting.Then set up the relationship model between parametric geometry topology constraints and numerical constraints according to the numerical and geometric constraints between the parameters.Then studied how to develop the secondary exploration of cheongsam intelligent model by using VBA language according to add-in program thus to set up cheongsam's parametric intelligent model.The fifth chapter verified the automatic generation of the cheongsam sample plate's availability from theoretical analysis and virtual fitting effect and the clothing affect these three aspects.The theoretical analysis tested the seamability of the plate,sample.The virtual fitting part used virtual fitting software CLO 3D to generate virtual fitting models according to the three-dimensional anthropometric data and to do the virtual stitching of two-dimensional cheongsam sample plate.The effect of stitching was evaluated through the three evaluation systems of the clothing perspective,the tension of clothing,clothing and human contact point.The effect of clothing mainly through the production of clothing and rely on expert evaluation method to conduct a subjective evaluation of the rationality of the model.