青年男性肩点横截面曲线及其体型细分  

Study on subdivision of young male's shoulder shapes and cross-section curve

作　　者：金娟凤[1] 庞程方[1] 陈伟杰[2] 叶晓露[1] 邹奉元[1,3] 

机构地区：[1]浙江理工大学服装学院,浙江杭州310018 [2]浙江理工大学机械与自动控制学院,浙江杭州310018 [3]浙江省服装工程技术研究中心,浙江杭州310018

出　　处：《纺织学报》2016年第8期100-106,共7页Journal of Textile Research

基　　金：国家国际科技合作专项项目(2011DFB51570)

摘　　要：为建立基于横截面曲线形态指标细分青年男性肩部体型的方法,选取了年龄在18～26周岁的213名在校青年男性,用三维人体扫描仪测量获得其肩部数据,借助IMAGEWARE 12.0与Mat Lab R2012b软件,利用冒泡排序算法实现了肩点的自动准确提取,且肩高相对误差控制在（0,0.011）之间;通过分析肩点横截面形态特征,提取了曲线曲率半径及矢额径比作为分类指标,对比曲率半径均值,确定了5个特征点角度为-50°、-35°、0°、50°与90°;应用K-Means聚类方法及方差分析,将肩部体型划分为4类,建立了依据特征点曲率半径与矢额径比细分肩部体型的方法,通过对比肩点截面曲线形态,进一步定性描述了4类肩部截面的曲线变化特征。In order to establish a new method of subdividing young male＇ s shoulder shapes based on morphology of cross-section curve,measurement of 213 young college male students aged from 18 to 26 years old were taken with a 3-D body scanner;using software tools of Imageware 12.0 and Mat Lab R2012 b,shoulder point was found out automatically using bubble sort algorithm,and relative error of shoulder height was managed to control between（0,1.1%）;curvature radius of feature points and the amount of sagittaldiameter-ratio were extracted through analyzing the morphology of shoulder point cross-section curve,and-50°,-35°,0°,50° and 90° were identified as five feature points by comparing the mean value of the radius of curvature;shoulder shape types were subdivided into 4 types by using K-Means clustering algorithm and analysis of variance,and a new method of somatotype was established based on the curvature radius of feature points and the amount of sagittal-diameter-ratio;finally,the variation characteristics of 4types was verified by comparing the sectional feature curves of each type of shoulder.

关 键 词：肩部 截面曲线 冒泡排序算法 曲率半径 矢额径比 聚类分析 体型细分 

分 类 号：TS941.17[轻工技术与工程—服装设计与工程]