To set the garment purchasing system using the network and cellular phone in IT, authors investigated 3-dimensioanl body (referred to as “3D-body” hereinafter) shape images for well-suited garment selection for adult women. In this study, authors extracted the evaluated information to show the relationship between the expression of body image key-words and the representation of the 3D-body modeling simulation. In order to select women's garments well-suited for 3D-body image, 72 subjects of the 3D-body model were adopted from a broad range of women ranging in age from the late 20's to 70's.
In this first part of the study, the five principal components of the 3D-body shape images were extracted based on evaluation scales with values from 1 to 5 for full-length 3D-body images Al to A6 and partial 3D-body images B 1 to B19 (number of evaluation respondents = 81) . The 3D-body shape image were classified into six clusters i.e., Ideal figure, Underweight figure, Overweight figure, Slightly masculine figure, and other image types, using those 4 principal components scores.

View full abstract

In this second part of our papers, for a 3D-body shape image information of a garment design selection support system in adult women, the categorized 3D-body shapes by the qualitative image evaluations using the interval scale in the part one is extracted from the size features using the 1D-body measurements of the quantitative ratio scales. The four 3D-body image factors (Overweight figure - Ideal figure within Underweight and Standard figure, Vertical length proportions images, Feminine-Masculine figure images, Shoulder size and shapes images) were able to be predicted using both the interval scale of qualitative 3D-body images and the quantitative ratio scale of the 1D-body measurements by means of the multiple regression analysis. However it was difficult to predict “Feminine-Masculine figure images”

View full abstract