3D Virtual Modeling Of Female Upper Body And Automatic Pattern Generation

With the rapid advancement of computer information technology,the traditional clothing industry has stridden towards automation and digitization that drive the growth of electronic commerce and line retailing.Now,the developing tendency of 3D garment CAD system mainly shows as follows: virtual sample manufacture(i.e.,tailor made),remote clothing fitting(i.e.,virtual fitting),virtual clothing design,garment pattern generation with quick response and so on.Many technologies need to be used to realize automatic generation of garment pattern,such as body measurement,modeling and pattern generation.Therefore,many researchers have focused on how to combine the three technologies according to the computer graphics,geometryand mechanics.This study proposed an approach on 3D upper body modeling and automatic pattern generation based on the body measurements extracted by non-contact anthropometry.Firstly,based on the frontal and side images of the human body,the automatic extraction system of body sizes was developed through silhouette extraction,identification of landmarks and girth prediction.Secondly,the generation rules of 15 characteristic cross-sectional curves were established using a method “feature points – inserted points – feature curves – basic surface – mannequin”,to develop a 3D virtual body modeling system that can fit the body type very well.Third,the relationship between the human body and pattern parameters was analyzed based on two types of women's suits,and then the pattern generation rules were built to realize the automation of individualized pattern generation according to the body characteristics.Finally,the mathematical models of ease distributions were built throughthe spatial distance ease between the body and garment surfaces,and the basic patterns were altered to get the final garment patterns by converting the distance ease into the pattern increments.These works were described in detail as follows:(1)Improvement of automatic body measuring technologyIn this study,two cameras were controlled to take the frontal and side digital images of the human body simultaneously to eliminate errors arising from body breathing and sway.Since the focal lengths and the shooting distances of the two cameras may differ slightly,an experiment of calibration with known targets was implemented to make sure their spatial resolutions(i.e.,pixels per mm)were identical.By considering the difference of body type,each position was classified into subcategories during girth prediction.Since the neck shape had four categories,the bust,under-bust,waist and abdomen shapes all had three categories,and the hip shape had two categories,the human body was classified into 648 subcategories.Therefore,the regression models were considered to be effective according to the verification,which showed more than 96% of the predicted data at under-bust,waist and abdomen were within the required tolerances for manufacturing dress and women's suits.(2)Generation rules of characteristic cross-sectional curvesSince the feature positions(e.x.,bust,waist and hip)which are related to garment patterns,are necessary for modeling virtual mannequin,the generation rules of characteristic cross-sectional curves are key to model the individualized female upper body.According to the method “feature points – inserted points – feature curves – basic surface – mannequin”,the cross-sectional curves of 15 characteristic positions were analyzed.The feature points of each position were determined at each curve,such as the side neck point,front neck point,shoulder point,bust point,and bust root point and so on to get the cross sections,and then some feature points were inserted at the curves according to the widths and depths to establish the calculative models.For example,there are 18 points distributed at the bust cross-sectional curve to determine the shape.The final mannequin could describe the basic characteristics of a human body,and the shape of the feature curves could also fit the body type to provide basis for the future research on automatic pattern generation.(3)Prototype pattern generation rules of women's suitsMost current patternmaking systems are actually semi-automatic,and there are only standard pattern database for some garment styles in these systems without the combination with the human body.In this study,each structuralpoint,line and curve of the garment pattern was analyzed to understand the relationship between the human body and the garment pattern.By converting the body measurements into the pattern parameters,the pattern generation rules of the women's suits were built based on the body sizes and the shape of the characteristic cross-sectional curves,to realize the automation of pattern generation.(4)Mathematical models of ease distributionTo realize the individualized apparel pattern generation,the ease allowance needs to be added into the prototype.Based on the spatial ease distribution,the distance eases between the body and garment surfaces were measured to establish the mathematical models of ease distribution.Then the distance ease was converted into the pattern increments by curve fitting on the virtual mannequin to realize the conversion from prototype to individualized apparel patterns.This study can realize 3D virtual modeling of female upper body and the automatic generation of the individualized apparel patterns based on the frontal and side images.This study proposed a new approachon pattern generation by converting the body measurements and body characteristics into pattern parameters,instead of relying on the traditional patternmaking experience or the computer fitting technology.