남성 정장 재킷의 대량맞춤생산을 위한 체형 분류 방법

Abstract

Background/Objectives: The purpose of present study is to develop a new body shape classification method, that can be used for the mass customization of men’s formal jackets, using variables measured on the men’s flattened body surfaces.
 Methods: Male participants in their 20s-50s were extracted from the 5th Size Korea database. The body surfaces were flattened into three-panels according to the conventional structure of men’s formal jackets. Variables were measured on the flattened panels, and those were inputted into the PCA. K-means clustering was conducted to categorize the flattened figure types. A logistic regression model was established to predict the body surface types.
 Results: 152 individuals were extracted for the body surface flattening. The 3D meshes were formed on the body surfaces, and those were flattened using distortion minimizing algorithm. Seventeen angles and two size differences were measured on the flattened figures, and five factors were extracted by the PCA. Five flattened figure types were extracted through k-means clustering: “normal”, “sway-back”, “bend-forward” “lean-back-b” and “lean-back-I”. The correspondence between the logistic regression model and K-means clustering was 90.79%.
 Conclusion/Implications: The classification method developed in the present study enable the quick and convenient decision of apparel customer's body shape type. It is optimized for mass customization of men's jackets because the variables were measured by converting the body surface into a jacket pattern structure.