基于青年女性下肢形态分类的特征部位围度拟合
来源:用户上传
作者:徐凯忆 钟泽君 蔡晓裕 顾冰菲
摘 要:为实现个性化裤装样板自动生成,探讨了青年女性下肢体型的分类方法及腰部、腹部、臀部、大腿根部和膝部5个人体特征部位的围度预测模型建立。主要通过美国[TC]2三维人体扫描仪获取202名在校女大学生的人体点云数据,测量各特征部位的围度、宽度和厚度等相关形态参数。然后进行体型分类,提出VDwh值(表征腰臀相对凸出量)、Dbw值(表征后腰的凹陷程度)、HDht值(表征站姿)、Atk值(表征腿型)4个形态指标,将青年女性下肢形态分为圆长型、圆润型、扁长型三类。基于体型分类结果,选择圆润体,对该体型下5个特征部位的宽度、厚度以及围度进行相关分析并建立特征部位围度的回归方程,并以手工数据进行验证分析。结果表明:除大腿根部最大误差值为1.98cm外,分类后预测的其他特征部位围度值与手工数据的误差绝对值均在1.5cm范围内,说明本围度预测方法具有较高的准确性,对基于照片的青年女体个性化裤装样板自动生成提供了一定的技术支撑。
关键词:人体测量;形态分类;围度预测;回归分析
中图分类号:TS 941.17
文献标志码:A
文章编号:1009-265X(2022)01-0204-08
Abstract: In order to achieve the automatic generation of personalized pant templates, this study discussed the classification method of morphologies of young women's lower limbs and set up a circumference prediction model at five characteristic parts of human body, that is, the waist, abdomen, hip, thigh and knee. The point cloud data of 202 female college students were obtained by the U.S. [TC]2 3D human body scanner, and relevant morphological parameters, i.e., circumference, width and thickness of each characteristic part were measured. Then the morphologies were divided into three types: oblong, round, and prolate, and four morphological indicators were proposed: VDwh(representing the relative protrusion of waist and hip), Dbw(representing the depression degree in the lower waist), HDht(representing the standing posture), and Atk(representing the leg type). Based on the classification results of morphologies, the morphology of "round" was selected to analyze the width, thickness and circumference of five characteristic parts under this morphology and a regression equation of the circumferences at the characteristic parts were established. The calculation models were verified through a comparison with the manual data. The results show that except that the maximum error at the thigh is 1.98cm, the absolute values of the errors between the predicted circumference values of characteristic parts and manual data are all within the range of 1.5cm, which shows that this circumference prediction method has high accuracy. The present study provides some technical support for the automatic generation of personalized pant templates for young women based on photos.
Key words: anthropometry; morphological classification; circumference prediction; regressionanalysis
S着信息化时代的普及,以手工制作为主要工艺的传统服装行业正迅速朝着自动化、数字化方向发展。对于服装行业来说,服装合体度日趋受到生产厂商的重视,逐渐成为提升消费者满意度的主要因素,在裤装纸样设计中,合体度尤为重要,而人体重要部位的围度大小是影响合体度舒适度的重要因素[1-2]。
为了满足消费者的需求,国内外学者们对人体的特征部位进行了体型细分。Gupta[3]选取18~26岁的大学生作为实验对象,测量他们身上29个测量项目,运用数理统计的方法,将这500个实验对象细分成5种体型;Yoon等[4]提出采用三维空间矢量方向角的方法对人体的上侧面体型进行合理的分类;Vuruskan等[5]将人体的胸围、腰围、臀围等作为基本参数,并采用主观视觉分析方法,得出5种不同的人体体型;金娟凤等[6]以肩部曲线的曲率半径和矢额径比为指标研究肩部的横截面形态,将人体的肩部体型细分成4类;王丽等[7]在提取乳间距、胸围、乳点高等20个与胸部有关的参数数据后,进行PCA降维处理,细分胸部的截面形态;王祺明[8]通过自编程序获取人体胸腰臀3个部位的截面面积与围度平方,并以两者的商作为特征变量,将女性体型细分为3类;姚怡等[9]基于人体前后中心线、胸点、背突点、臀突点的矢状面曲线以及人体肩部、侧缝处冠状面曲线对人体纵截面体型进行分类。目前学者在体型分类方面的研究多集中在人体躯干部分,但关于人体下肢形态分类的研究较少,而人体下肢局部特征及其分类研究是裤装结构设计、舒适性、合体性研究及版型优化的重要基础,因此对下肢的研究也显得尤为重要。
转载注明来源:https://www.xzbu.com/1/view-15422768.htm