纺织学报 ›› 2019, Vol. 40 ›› Issue (05): 124-130.doi: 10.13475/j.fzxb.20180607907

• 服装工程 • 上一篇    下一篇

基于改进层次分析法的特殊体型样板识别

周捷1(), 李健1, 马秋瑞1, 黄晓杰2   

  1. 1.西安工程大学 服装与艺术设计学院, 陕西 西安 710048
    2.中原工学院 信息商务学院, 河南 郑州 450007
  • 收稿日期:2018-06-25 修回日期:2019-01-29 出版日期:2019-05-15 发布日期:2019-05-21
  • 作者简介:周捷(1969—),女,副教授,博士。主要研究方向为功能性内衣、服装结构和人体科学等。E-mail: xianzj99@163.com。
  • 基金资助:
    陕西省科技厅国际科技合作计划项目(2018KW-056);陕西高等教育教改研究项目(17BZ037)

Recognition of special template based on improved analytic hierarchy process

ZHOU Jie1(), LI Jian1, MA Qiurui1, HUANG Xiaojie2   

  1. 1. School of Apparel and Art Design, Xi'an Polytechnic University, Xi'an, Shaanxi 710048, China
    2. College of Information & Business, Zhongyuan University of Technology, Zhengzhou, Henan 450007, China
  • Received:2018-06-25 Revised:2019-01-29 Online:2019-05-15 Published:2019-05-21

RichHTML

6

PDF (PC)

370

摘要：

针对传统层次分析法的缺陷,提出结合因子分析、聚类分析的改进层次分析方法,应用到特体样板识别中。首先,通过因子分析确定各级指标的影响因素及其权重系数。然后,采用K-means聚类分析将特体样板聚为臀部、胸部和腹部等3类典型样板。最后,运用层次分析法确定各指标的权重,构建了包括6个二级指标、13个三级指标、3个四级指标的特体样板递阶层次结构模型。随机选取3个样本进行实证研究,以因子分析中提取的5个主成分为聚类指标,以K-means方法进行聚类,采用特体样板递阶层次结构模型识别被测样本与臀部、胸部、腹部样板的隶属度。结果表明,该方法可有效地表征特体样板的变异程度并识别所属的样板类别。

关键词: 特体样板, 因子分析, 聚类分析, 改进层次分析法

Abstract:

Aiming at the shortcomings of conventional analytic hierarchy process, an improved analytic hierarchy process based on combination factor analysis and cluster analysis was proposed and applied to special model recognition. Firstly, factor analysis was applied to determine the influencing factors and weight coefficients of each level of indicators. Then, K-means cluster analysis was applied to collect the special samples into three typical models such as hip, chest and abdomen. Finally, the analytic hierarchy process was applied to determine the weight of each indicator, and a special model-level hierarchical structure model including six secondary indicators, 13 third-level indicators and three four-level indicators was constructed. Three samples were randomly selected for empirical research. The five principal components extracted from the factor analysis were used as clustering indicators, and the K-means method was used for clustering. The specific sample-level hierarchical structure model was applied to identify the sample to be tested and hip and chest. The results show that the method can effectively characterize the variation of the special model and identify the model category.

Key words: special template, factor analysis, cluster analysis, improved analytic hierarchy process

中图分类号: 

  • TS941.712

图1

特体修正样板的典型指标"

图2

改进层次分析法流程图"

表1

主成分因子分析"

成分 初始特征值 因子旋转后
特征根 方差贡
献率/%		 累计贡
献率/%		 特征根 方差贡
献率/%		 累计贡
献率/%
1 3.232 24.865 24.865 2.740 21.076 21.076
2 2.281 17.545 42.410 2.311 17.777 38.853
3 1.882 14.473 56.884 2.189 16.840 55.693
4 1.139 8.764 65.648 1.274 9.800 65.493
5 1.027 7.902 73.550 1.047 8.057 73.550

表2

旋转载荷矩阵"

评价
指标		 成分
1 2 3 4 5
C1 -0.004 -0.117 0.898 0.019 -0.053
C2 0.826 -0.133 0.075 -0.010 0.124
C3 0.557 0.241 0.017 -0.267 0.116
C4 0.095 0.537 -0.296 -0.062 -0.531
C5 0.734 0.202 -0.003 0.489 -0.190
C6 0.793 0.083 -0.148 -0.132 0.011
C7 -0.035 -0.087 0.716 0.123 0.141
C8 0.068 0.952 -0.035 0.032 0.102
C9 0.742 0.045 -0.069 0.501 -0.194
C10 -0.031 0.090 0.854 -0.063 -0.114
C11 0.073 0.253 -0.106 0.055 0.765
C12 -0.061 0.007 0.081 0.817 0.126
C13 -0.075 -0.947 0.007 -0.024 -0.128

表3

方差分析表"

聚类数 变量 类间
均方		 误差
均方		 F值 Sig.
3 因子1 35.603 0.810 43.960 0.000
因子2 96.744 0.474 204.132 0.000
因子3 65.601 0.645 101.699 0.000
因子4 6.849 0.968 7.077 0.001
因子5 8.209 0.960 8.547 0.000
4 因子1 16.043 0.876 18.321 0.000
因子2 45.592 0.631 72.200 0.000
因子3 38.712 0.688 56.241 0.000
因子4 60.413 0.509 118.692 0.000
因子5 37.000 0.702 52.671 0.000
5 因子1 35.831 0.615 58.250 0.000
因子2 50.323 0.455 110.602 0.000
因子3 40.349 0.565 71.390 0.000
因子4 49.733 0.462 107.758 0.000
因子5 15.047 0.845 17.811 0.000

表4

3类特体样板主要部位平均修正量统计表"

样板类型 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13
D1 7.72 8.95 -2.32 -7.94 8.96 11.56 1.27 -2.43 7.41 -1.39 -1.85 0.17 2.00
D2 -0.85 -6.17 -8.75 -12.13 -9.95 -3.68 -2.42 -13.35 -1.95 -10.43 -9.37 -0.30 13.02
D3 -32.73 -5.28 -12.13 -2.09 3.55 4.60 -6.51 -0.78 4.68 -15.28 -2.28 0.61 0.76

图3

特体样板识别模型的递阶层次结构"

表5

判断矩阵标度aij含义"

标度(aij ) 含义
1 元素i与元素j同等重要
3 元素i比元素j稍重要
5 元素i比元素j颇重要
7 元素i比元素j极重要
9 元素i比元素j极端重要
2/4/6/8 分别介于以上标度的中间值
倒数 若元素j与元素i的关系是: aji=1/aij

表6

特体样板识别判断矩阵B-A"

类别 B1 B2 B3 B4 B5 权重
B1 1 7/6 21/17 7/3 21/8 0.290 0
B2 6/7 1 15/14 19/11 19/9 0.238 8
B3 17/21 14/15 1 19/11 2 0.227 2
B4 3/7 11/19 11/19 1 5/4 0.133 1
B5 8/21 9/19 1/2 4/5 1 0.110 9

表7

廓形因子判断矩阵C-B1"

类别1 C2 C3 C5 C6 C9 权重
C2 1 7 5 3 4 0.484 5
C3 1/7 1 1/3 1/5 1/4 0.043 4
C5 1/5 3 1 1/3 1/2 0.089 8
C6 1/3 5 3 1 3 0.249 4
C9 1/4 4 2 1/3 1 0.132 9

表8

待测样本与特体样板主要部位尺寸差值大小"

部位
指标		 修正量 样本a与3类特体样板差值 样本b与3类特体样板差值 样本c与3类特体样板差值
样本a 样本b 样本c Δa1 Δa2 Δa3 Δb1 Δb2 Δb3 Δc1 Δc2 Δc3
C1 30 70 -70 22.28 30.85 62.73 62.28 70.85 102.73 -77.72 -69.15 -37.27
C2 -30 60 -40 -38.95 -23.83 -24.72 51.05 66.17 65.28 -48.95 -33.83 -34.72
C3 -15 0 -15 -12.68 -6.25 -9.38 2.32 8.75 5.62 -12.68 -6.25 -9.38
C4 -12 -20 -15 -4.06 0.13 -9.91 -12.06 -7.87 -17.91 -7.06 -2.87 -12.91
C5 8 45 -60 -0.96 17.95 4.45 36.04 54.95 41.45 -68.96 -50.05 -63.55
C6 -15 10 -28 -26.56 -11.32 -19.6 -1.56 13.68 5.4 -39.56 -24.32 -32.6
C7 -10 15 -15 -11.27 -7.58 -3.49 13.73 17.42 21.51 -16.27 -12.58 -8.49
C8 0 0 -10 2.43 13.35 0.78 2.43 13.35 0.78 -7.57 3.35 -9.22
C9 3 45 -40 -4.41 4.95 -1.68 37.59 46.95 40.32 -47.41 -38.05 -44.68
C10 10 35 -35 11.39 20.43 25.82 36.39 45.43 50.82 -33.61 -24.57 -19.18
C11 0 5 0 1.85 9.37 2.28 6.85 14.37 7.28 1.85 9.37 2.28
C12 10 0 0 9.83 10.3 9.39 -0.17 0.3 -0.61 -0.17 0.3 -0.61
C13 0 0 10 -2 -13.02 -0.76 -2 -13.02 -0.76 8 -3.02 9.24

表9

指标层总权重计算和排序"

二级指标 三级指标 四级指标单
排序权重
WD		 总排序权重
W=WB×
WC×WD
指标 权重
WB		 指标 单排序
权重WC
B1 0.290 0 C2 0.484 5 [0.700 7,0.097 2,
0.202 1]		 W1=
0.293 7,
W2=
0.436 4,
W3=
0.269 9
C3 0.043 4 [0.637 0, 0.104 7,
0.258 3]
C5 0.089 8 [0.058 1, 0.735 2,
0.206 7]
C6 0.249 4 [0.637 0, 0.104 7,
0.258 3]
C9 0.132 9 [0.296 9, 0.617 5,
0.085 6]
B2 0.238 8 C8 0.750 0 [0.218 5, 0.714 7,
0.066 8]
C13 0.250 0 [0.148 8, 0.785 4,
0.065 8]
B3 0.227 2 C1 0.637 0 [0.081 0, 0.188 4,
0.730 6]
C7 0.104 7 [0.637 0, 0.258 3,
0.104 7]
C10 0.258 3 [0.071 9, 0.279 0,
0.649 1]
B4 0.133 1 C12 1.000 [0.280 8, 0.584 2,
0.135 0]
B5 0.110 9 C4 0.166 7 [0.227 1, 0.051 0,
0.721 9]
C11 0.833 3 [0.081 0, 0.730 6,
0.188 4]

图4

特体样板类别雷达图"

[1] CORONA E, ALENYA G, GABAS A, et al. Active garment recognition and target grasping point detection using deep learning[J]. Pattern Recognition, 2018,74(2):629-641.
[2] APEAGYEI P R. Application of 3D body scanning technology to human measurement for clothing fit[J]. International Journal of Digital Content Technology & Its Applications, 2010,4(7):58-68.
[3] MENG Y, MOK P Y, JIN X. Computer aided clothing pattern design with 3D editing and pattern alteration[J]. Computer Aided Design, 2012,44(8):721-734.
[4] 冀艳波, 张睿, 应柏安. 男西装工业生产用服装样板改进分析[J]. 西安工程大学学报, 2008,22(5):565-568.
JI Yanbo, ZHANG Rui, YING Boan. Analysis of the improved industry production pattern of the man suit[J]. Journal of Xi'an Polytechnic University, 2008,22(5):565-568.
[5] 张伶俐, 张皋鹏. 应用MatLab的服装纸样参数化平面制版[J]. 纺织学报, 2019,40(1):130-135.
ZHANG Linli, ZHANG Gaopeng. Parametric flat pattern design for clothing based on MatLab[J]. Journal of Textile Research, 2019,40(1):130-135.
[6] 朱伟明, 谢琴, 彭卉. 男西服数字化智能化量身定制系统研发[J]. 纺织学报, 2017,38(04):151-157.
ZHU Weiming, XIE Qin, PENG Hui. Digitalize and intelligentize customization system of men's suit[J]. Journal of Textile Research, 2017,38(04):151-157.
[7] 彭磊, 谢红, 邹奇芝. 基于量身定制特体女装原型样板的生成[J]. 纺织学报, 2011,32(4):101-105.
PENG Lei, XIE Hong, ZOU Qizhi. Generation of women's basic patterns of special body forms based on made to measure[J]. Journal of Textile Research, 2011,32(4):101-105.
[8] LIU K, WANG J, KAMALHA E, et al. Construction of a prediction model for body dimensions used in garment pattern making based on anthropometric data learning[J]. The Journal of The Textile Institute, 2017,108(12):2107-2114.
[9] 刘为敏, 谢红. BP神经网络下的智能化合体服装样板生成[J]. 纺织学报, 2018,39(7):116-121.
LIU Weimin, XIE Hong. Generation of intelligent fitting pattern based on BP neural network[J]. Journal of Textile Research, 2018,39(7):116-121.
[10] PENG S, LIU Z W, The comprehensive assessment of power quality based on attribute recognition theory and AHP algorithm[J]. Advanced Materials Research, 2014, 1044-1045:515-522.
[11] PUNIASINDHU S, NEHRAV, LUTHRA S. Recognition and prioritization of challenges in growth of solar energy using analytical hierarchy process: Indian outlook[J]. Energy, 2016,100(4):332-348.
[12] ZHAO J, JIN J, ZHU J, et al. Water resources risk assessment model based on the subjective and objective Combination weighting methods[J]. Water Resources Management, 2016,30(9):3027-3042.
[13] SAATY T L, VARGASL G. An innovative orders-of-magnitude approach to AHP-based mutli-criteria decision making: prioritizing divergent intangible humane acts[J]. European Journal of Operational Research, 2011,214(3):703-715.
[14] 郑唯唯, 凌晓静, 杨萍. 基于层次分析法的多指标灰靶决策模型[J]. 纺织高校基础科学学报, 2013(4):463-467.
ZHENG Weiwei, LING Xiaojing, YANG Ping. The multi-attribute grey target decision-making model based on analytic hierarchy process[J]. Basic Sciences Journal of Textile Universities, 2013 (4):463-467.
[15] 陈明月, 杨青, 高琼琼. 层次分析法在本科教学评估中的应用研究[J]. 西安工程大学学报, 2010,24(1):106-110.
CHEN Mingyue, YANG Qing, GAO Qiongqiong. Application of AHP in the undergraduate education evaluate[J]. Journal of Xi'an Polytechnic University, 2010,24(1):106-110.
[16] OKABE K, KUROKAWA T. Silhouette analysis and categorization of body shape changes in Japanese girls during periods of growth[J]. Japanese Journal of Ergonomics, 2010,41(1):17-23.
[17] 齐静, 李毅, 张欣. 我国西部地区青年男性体型描述与体型分类研究[J]. 纺织学报, 2010,31(5):107-111.
QI Jing, LI Yi, ZHANG Xin. Description and classification of young men's body type in the west China[J]. Journal of Textile Research, 2010,31(5):107-111.
[18] 庹武, 郑攀, 常亭亭, 等. 基于层次分析法的服装裁剪分床影响因素[J]. 纺织学报, 2013,34(4):148-152.
TUO Wu, ZHENG Pan, CHANG Tingting, et al. Analysis of factors affecting cutting scheme based on analytical hierarchy process[J]. Journal of Textile Research, 2013,34(4):148-152.
[19] ZHANG J, WANG C M, LIU L, et al. Investigation of carbon dioxide emission in China by primary component analysis[J]. Science of the Total Environment, 2014,472(2):239-247.
[20] WANG R W, YAN J L, WU X Y, et al. Automatic Identification of lyocell and cotton fibers using cluster Analysis[J]. Textile Research Journal, 2010,80(13):1330-1334.
[1] 邢丽娟, 刘新金, 苏旭中, 曹秀明. 应用灰色聚类方法评价特种动物纤维综合物理性能[J]. 纺织学报, 2019, 40(01): 26-31.
[2] 高广明 朱江晖 许丽慧 孙莉 谷芬芬. C2C环境下服装零售感知服务质量综合评价模型[J]. 纺织学报, 2018, 39(08): 164-170.
[3] 王军 李晓久 潘力 姚彤 于佐君. 东北地区青年女性腰臀部体型特征与分类[J]. 纺织学报, 2018, 39(04): 106-110.
[4] 杜玉红 杨程午 蒋秀明 刘仁杰 蔡文超. 应用聚类神经网络的异纤检测多类光源优化设计[J]. 纺织学报, 2017, 38(10): 104-112.
[5] 王晓予 向军 潘如如 梁惠娥 高卫东. 服饰刺绣图案的自动提取与色块分割[J]. 纺织学报, 2017, 38(09): 120-126.
[6] 夏凤勤 毋戈 谢昊洋 钟跃崎. 基于人体纵截面特征曲线的体型分类[J]. 纺织学报, 2017, 38(06): 86-91.
[7] 高广明 朱江晖 谷芬芬. 服装卖场顾客感知服务质量评价体系与模型[J]. 纺织学报, 2017, 38(06): 136-142.
[8] 朱江晖 朱俊莉. 服装品牌形象影响因素及其权重分布[J]. 纺织学报, 2016, 37(4): 148-152.
[9] 黄帅 张毅 周志华. 采用因子分析法的服用织物电磁屏蔽性能影响因素分析[J]. 纺织学报, 2016, 37(2): 149-154.
[10] 马崇启 蔡薇琦 阚永葭. 酚醛纤维织物热湿舒适性的灰色聚类分析[J]. 纺织学报, 2016, 37(12): 29-32.
[11] 高融 王玲 杨以雄. 服装线上零售渠道消费者决策机制[J]. 纺织学报, 2016, 37(12): 146-152.
[12] 蔡薇琦 马崇启 阚永葭 杨金莲 李君丽 . 灰色聚类分析在织物热学性能评价中的应用[J]. 纺织学报, 2016, 37(11): 64-67.
[13] 周小溪 梁惠娥 陈潇潇 董稚雅 秦潇璇. 春夏季衬衫用色织面料材质的感觉评价[J]. 纺织学报, 2016, 37(08): 59-64.
[14] 金娟凤 庞程方 陈伟杰 叶晓露 邹奉元. 青年男性肩点横截面曲线及其体型细分[J]. 纺织学报, 2016, 37(08): 100-106.
[15] 周小溪 梁惠娥. 基于感性意象的服装面料设计效果评价[J]. 纺织学报, 2015, 36(04): 60-64.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
京ICP备14006648号　 京公网安备11010502044800号
版权所有 © 2021 《纺织学报》编辑部
地址： 北京市朝阳区延静里中街3号主楼6层《纺织学报》杂志社　邮政编码：100025　
电话：010-65017711,65917740　传真：010-65016539　Email:fangzhixuebao@vip.126.com
本系统由北京玛格泰克科技发展有限公司设计开发