基于纵向轮廓曲线的青年女性体型细分研究  

Research of young female somatotype based on longitudinal profile curves

作　　者：倪世明[1] 金娟凤[1] 庞程方[1] 邹奉元[1,2] 

机构地区：[1]浙江理工大学服装学院,浙江杭州310018 [2]浙江理工大学浙江省服装工程技术研究中心,浙江杭州310018

出　　处：《纺织学报》2014年第8期87-93,共7页Journal of Textile Research

基　　金：浙江省自然科学基金资助项目(Y1110504);国家国际科技合作专项项目(2011DFB51570);浙江理工大学研究生创新研究项目(YCX13016)

摘　　要：提取了能用于青年女性体型分类的纵向轮廓曲线,包括矢状面和冠状面轮廓曲线。建立了用轮廓曲线特征点曲率半径进行青年女性体型分类的方法。在国标体型分类的基础上,进一步细分青年女性体型,从纵向轮廓曲线形态入手,分析曲线特征点曲率半径,应用K-means聚类算法进行动态聚类,将伪F统计量作为判别函数来确定最佳类数目,最终将纵向体型分为8类,量化了青年女性纵向体型差异,并提出一种新的体型标识"国标体型+纵向体型"。This paper extracted the longitudinal profile curves which could be used for young female somatotype, including sagittal and coronal contour curves. A method of young woman shape classification was established based on the curvature radius of contour feature points. On the basis of GBbody shape classification, subdividing of young female body shape was conducted further, analyzing the characteristic of the curve radius of curvature from the perspective of the longitudinal profile curve shape. K-means clustering algorithm was used for dynamic clustering, and the pseudo F statistics were regarded as a discriminant function to determine the optimal class number. The final longitudinal shape could be divided into 8 classes, quantitatively distinguishing the young female longitudinal size differences. And a new shape identification of ＂national standard body shape ＋ longitudinal body shape＂ is put forward.

关 键 词：矢状面 冠状面 曲率半径 K-MEANS聚类 伪F统计量 体型细分 

分 类 号：TS941.17[轻工技术与工程—服装设计与工程]