摘要
为确定青年职场女性正面体型划分依据,采集了208位25~35岁年龄段青年职场女性身着紧身瑜伽服的正面照片,提取人体肩、腰、髋等部位横向投影宽度及肩高、腰高差等纵向投影高度,通过主成分分析及聚类分析,得到女性正面体型的分类结果,从中提取分类规则并验证。结果表明:基于肩髋宽度比、肩腰宽度比、腰髋宽度比及大腿根髋宽度比4个变量建立的女性正面体型的分类规则可以将正面体型分为X型、A型、H型以及O型4类。样本检验准确率达到94.5%,说明了本文分类规则的有效性。该规则可为基于拍照的体型分类及基于体型的服装搭配与推荐系统的研发提供一定的参考。
In order to determine the basis for the division of young working women’s frontal body shape,the frontal photos of 208 young working women aged 25~35 wearing tight yoga clothes were collected,and the transverse projection width,shoulder height,waist height difference and other longitudinal projection heights of human shoulder,waist and hip were extracted.The classification results of women's frontal body shape were obtained through principal component analysis and cluster analysis,from which the classification rules were extracted and verified.The results show that the classification rules of female frontal body type established on the basis of four variables of waist width,shoulder waist width ratio,waist hip width ratio and thigh root hip width ratio can divide the frontal body shape into four types:X type,A type,H type and O type.The accuracy of sample test is 94.5%,which indicates the validity of the classification rules.The rules can provide some reference for body type classification based on photos and the research and development of clothing matching and recommendation system based on body type.
作者
杨玲
陈郁
YANG Ling;CHEN Yu(College of Fashion and Textiles,Shanghai University of Engineering Science,Shanghai,201600,China)
出处
《浙江纺织服装职业技术学院学报》
2023年第2期22-27,共6页
Journal of Zhejiang Fashion Institute of Technology
基金
上海市教育委员会东方学者项目(编号:TP2017074)。
关键词
正面体型
体型分类
二维图像
聚类分析
主成分分析
frontal body shape
body type classification
two dimensional images
cluster analysis
principal component analysis