| Body shape is the basis of clothing design and related fashion products research.In order to find the classification method of female body shape for online garment customization,407 young women aged 20-35 were anthropometric measured by combining 3D measurement and manual measurement.Seven main components related to human body shape were extracted by using factor analysis,including fullness,height,leg,lower limb width,trunk,shoulder and anterior axillary point.By using the correlation analysis,the body size characteristic index and the body shape characteristic index were extracted from the seven main components.In addition,to classify a body in a way that would be effective,in this study,the way of clustering in two steps was adopted.Start with the underbust flatness ratio as the characteristic index,the human body were divided into mormal,round and flat.Then,the secondary body types of the three types were divided into14 types with the ratio of height and waist,chest-waist difference,hip-waist difference as the characteristic index.In order to make the body shape classification system connect with the online garment customization system better,the ANN was used to establish the body shape discrimination model.Finally,the clothing size system was established and the prototype of the system was carried out.Through the evaluation,the method of body shape classification can be applied to the system well. |