I.Introduction
In the 21st century, the world has been exposed to global markets without boundaries. For an apparel industry to be competitive in such an environment, studies and data on the figure type and size of various human races are required.
A number of domestic and overseas projects, such as the launch of a spaceship in 2003, distributive liberalization in 2005, and the Beijing Olympics in 2008, have seen China gain the attention of the world as a large consumer market. The inflow of varied information, rise in fashion standards, and intellectual elevation of socioeconomic activity on the development of mass media and high educational backgrounds mean that Chinese consumers are increasingly interested in imperial products and improved quality in their fashion surroundings (Cha, 2008). A preference for fashion products, the actual condition of wearing and purchasing, and information about Chinese body types have become necessary to develop clothing goods that cater to such consumer requests.
Human body measurements are important for an exact analysis of body figure type, and obviously lead to advances in quality of fit. In the past, the length of the human body and its circumference were obtained using a measuring tape, yielding 2D data from 3D measurements. These data provide information on body size in a static situation, but do not provide 3D data on body shape. Two-dimensional methods are inaccurate because the 3D data differ for each person and circumstance. We have thus come to use a 3D whole body scanner for body measurements. Such scanners obtain a 3D figure, providing measurements for the body size and appearance without direct body measurements. In addition, highly authentic sampling information can be used repeatedly. This removes some of the obstacles to human body measurement (Adriana & Ashdown, 2008; Cha, 2012; Choi, 2008; Choi & Ashdown, 2011). In this study, we analyzed the figure type and lower-body size of Chinese adult females in their early twenties using data obtained from 3D whole body scanners. We also provide concrete information related to Chinese women’s lower bodies in terms of crotch width, crotch length, inseam, outseam, slope of waist, etc. that may be useful to the Chinese apparel industry when designing trouser pants for adult women.
II.Background
1Special features and types of 3D body scanner
Depending on their measurement method, 3D body scanners can be classified as laser, phase shift, infrared, or structured light projection scanners. The laser and infra-red methods gather data of the human body shape in the x-, y-, and z-planes after taking a photograph of the human body surface. These approaches allow a whole body scan, but the equipment is generally very large. The method of structured light projecting obtains the surface shape using cross stripes. This can cause some difficulties in obtaining an image because of the need to adjust the angle at each point. The phase shift technique reflects a separated pattern at the surface. This image is fed to a computer by a camera, and the phase shift method then calculates each pixel and provides the size. Although phase shift techniques can view the human body as several pieces, it is not easy to change a pixel into a surface. Various types of 3D body scanners are used around the world. The most widely used are the Whole Body Scanner WB4 (Cyberware Corporation), Vitus 3D Body Scanner (Techmath Corporation), Body Scanner (TC2 Corporation), and Triform Body Scan (Wicks & Wilson Corporation). The software employed in these machines includes DigiSize (Cyberware Corporation), Scanworx and Polyworx (Techmath Corporation), and 3D Body Measurement Soft System (TC2 Corporation) (Cha, 2012; Choi, 2008; Kim, 2006).
3D body measurements give exact data within a short time, and can continue to perform measurements over a given period (Lee et al., 2012; Song & Ashdown, 2011). However, it is difficult to obtain data for parts of the body that are covered or overlap, such as the armpit and crotch. In the case of direct body measurements, data will differ according to the measurer, environment, frequency, and so on. It also requires a lot of time for the measurement, but has the merit of providing data for covered areas, such as the crotch and armpit. 3D body scanners can obtain data on the whole body within a minute, including data on body size, shape, and posture.
2.Classification of the lower body figure
Body measurement is a basic method for obtaining characteristics of body size and shape. Most of all, an understanding of body figure is important for making ergonomic apparel that has a good fit (Adriana & Ashdown, 2008). The body measurements must analyze and systematize the body figure in connection with clothes. That is, they should classify body figures in terms of their relationship with relevant items of clothing.
Park (1998) studied the body measurements of 364 women in their twenties. This study performed a factor analysis and cluster analysis on the measurements. The factor analysis found a total of 10 distinct factors. Three clusters were categorized: cluster 1 had a flat and fat shape, cluster 2 had a big curved silhouette at the front, and cluster 3 had a normal figure.
Hong and Park (2003) measured 195 female students, aged from 18 to 23 years old. They also carried out a factor analysis and cluster analysis for a total of 38 items. Six factors and three discrete clusters were found.
Ha and Sung (2005) measured 138 females aged from their twenties to sixties. Each female satisfied three conditions: a 90 cm bust circumference, Rohrer index of 1.5, and relative weight of over 110%. The authors analyzed factors and clusters for a total of 28 items, and determined seven factors and four clusters.
Park (2006) measured 341 females, aged from 19 to 24 years old, in Shanghai and Beijing, China, and analyzed factors and clusters for a total of six items. The factor analysis found six factors, and the cluster analysis determined three distinct clusters.
Choi (2008) analyzed the factors and clusters given by a total of 45 items from the measurements of 1036 Korean females aged in their twenties and thirties. The factor analysis classified seven elements, and the cluster analysis determined one cluster with a thick figure with short legs, one with long legs and a thick waist, and a third cluster was thin and tall with long legs.
Song and Ashdown’s (2011) lower body shape categorization method was a multidimensional body shape analysis using a large anthropometric data set (SizeUSA) and statistical methods. This body shape analysis used measurement data of 2,981 women aged 18－35 years from the SizeUSA data set. A total of 15 drop values calculated as differences between measurements and 1 angle were used for the principal component analysis, including drop values for girth, back arc, front arc, width, depth, backs depth, and front depth. The factor analysis classified five elements. From cluster analysis, they identified three body shape groups curvy shape (curviest silhouette between waist level and hip level and the most prominent abdomen silhouette), hip tilt shape (most prominent buttocks and lower body that is tilted toward the back), and straight shape (noncurvy silhouette and less prominent buttocks).
3.Preceding studies about Chinese adult women’s body figure
Kim (2005) compared and analyzed partial proportion of body shape of Korean adult women with that of Chinese 19-50 aged adult women who live in Shanghai and Beijing, China, which the data of Chinese were from self-measurement. The result showed that Chinese women have longer lower body, higher hips, and more narrow width of waist compared to Korean women. Also, hips thickness of Chinese women is similar with that of Korean women, but Chinese women have thick breast, greatly grown around breast, small waist size, and small hip circumference.
Choi (2002) compared body shape of women of Korea and China using measurement data of 410 women, 210 Korean female college students and 200 Chinese female college students. Chinese female college students have big tilt of the shoulders and hips, and Korean female college students have healthy lower body, long legs and arms for the height, and large hips, compared to each country.
According to the study of Korean Apparel Industry Association (2004), Chinese adult women are taller than Korean adult women, but weight of those two are alike. So, it shows that Korean adult women are fat relatively. Chinese adult women have western body shape which has large breast and hips, small gourd-shaped waist, and very high waist height.
In the research of Samsung Fashion Institute (2003), it shows that the value of circumference and thickness are high in Chinese women, but the value of width is high in Korean women. So, it can be known that Chinese women show round shape with high flatness ratio relative to Korean women. Also, Chinese women have large breast and long upper body, but size of mammilla was small. In case of lower body, Korean women were thick especially in hip parts. In brief, Chinese women compared to Korean women have grown upper body especially in breast, but their shoulder were narrow, and in lower body, it is flat and total crotch length and hip length are long.
To summarize the above, in lower body, the body shape Chinese adult women are that circumference of waist and hips are small but tilt of hips is large, and total crotch length and hip height are long. In addition, difference in size of waist and hips is large and they have high waist height like western body shape. However, most pre-researches were through self-measurement and almost no studies were done with 3D measurement. Also, since most researches were concentrated in the upper body and overall proportion, there is not enough information on lower body shape of Chinese women. So, it needs to be studied deeply. Thus, this research tried to help in design of pants for Chinese adult women by analyzing lower body shape of Chinese adult women using data from 3D measurement.
III.Methodology
1.Subjects
A total of 210 women from University in China, were selected by simple random sampling. The women were aged from 18 to 24 years old. Informed consent was obtained from participants to report their body shapes and size data.
2.Period and place of measurement
A preliminary measuring test was conducted from 21－24 January, 2008, in the Human Engineering room in University in China. The main measurement was carried out from 20 February to 30 March, 2008.
3.Posture of 3D body measurement
The subjects were asked to stand with their feet 30 cm apart and their legs straight. To prevent the loss of data from the crotch and armpit areas, they were also required to hold handles and look forward (Cha, 2008; Choi & Ashdown, 2011) (Fig. 1).
4.Measurement instruments and garments
The 3D Whole Body Scanner (TC2 Corporation) was used for measurements (Fig. 2). The subjects wore only a thin pair of briefs. To prevent the loss of data from the crotch and armpit areas, they were also required to hold handles and look forward briefs without upper wear to ensure they were physically comfortable for the test, and wore a hat on their head after tying their hair.
5.Measurement items and analysis of 3D data
These data were analyzed using TC2 Corporation’s analysis program at the Hong Kong Polytechnic University. A total of 39 items were determined using the underbody auto-extract program. These opened and analyzed an Excel program after conversion from ORD files. The 39 items were as <Table 1> and <Fig. 3>. The DC Suite program of the Digital Clothing Center at Seoul National University was used to make 3D forms.
6.Analysis program and methods
The data obtained were processed using the SPSSWIN 15.0 software. The average and standard deviation of the measurements were calculated to determine the lower body characteristics of Chinese adult women, and an exploratory factor analysis was used to draw measurement item factors of the lower body. A nonhierarchical K-means cluster analysis was performed on the basis of the factor loadings for lower-body grouping. A one-way ANOVA test was then used to verify the difference between lower-body measurement’s items derived from the clusters of lower-body measurements. Post-verification was completed using Duncan’s multiple range tests. The statistical significance was examined at a level of α=0.05.
IV.Results
We obtained somatometric results by analyzing the size and shape of the lower half of the body using the 3D full body scanner. The minimum, maximum, mean, and standard deviation of 39 metrics are shown in <Table 2>.
Principal component analysis was used to reduce the number of variables, and an orthogonal rotation technique, Varimax, was used for factor rotation. Only variables with a factor loading of above 0.5 were considered meaningful. This analysis extracted five factors whose eigenvalue was greater than 1, and the total variance explanation power was found to be about 84.5%. The eigenvalue of factor 1 was approximately 16.8, and this factor explains 43.2% of the whole variance. We named this the ‘factor of lower-half obesity,’ because it consists of the waist circumference, hip circumference, thigh circumference, knee circumference, calf circumference, waist width, and hip width. The eigenvalue of factor 2 was about 8.6, and this factor explains approximately 22.1% of the whole variance. This was named the ‘factor of lower-half height,’ because it is composed of height, waist height, hip height, thigh height, abdomen height, calf height, outside leg height, inside leg height, and so on. Statistics for the other three factors, as well as a breakdown of the factor loadage, are given in <Table 3> and <Table 4>.
A cluster analysis was implemented to classify the lower-body types in terms of the factors determined from the factor analysis. A nonhierarchical K-Means clustering was used for cluster analysis, resulting in three appropriate groups to classify the lower-body types. Cluster 1 focuses on the lower-half body shape (n=79), having the highest values of obesity, especially the highest value in most of the circumference, width, and weight measurements and the lowest value of lower-half height. Cluster 2 has a small hip circumference and thin thigh compared to the waist and abdomen circumference, but exhibits the lowest values for hip circumference, hip width, abdomen height, thigh circumference, and weight, and middle values for most other categories. Cluster 3 contains high values of both lower-half height and length from waist to crotch (n=46). Therefore, the clusters were named obesity (cluster 1), small hip (cluster 2), and slim body with long legs (cluster 3) (Table 5).
Statistics for the average measurement size of each cluster are shown in <Table 6>. <Fig. 4> illustrate each of the three body types. For making avatar, using size is like this. First, ‘Obesity’ avatar has 77.7cm in waist circumference, waist height 95.9cm, hip circumference 75.5cm, waist to hip length 21cm, and knee height 43.1cm. Second, ‘Small hip’ has waist circumference 71.7cm, waist height 95.9cm, hip circumference 91.6cm, hip height 75.3cm, waist to hip length 21.4cm, and knee height 43.1cm. Third, ‘Slim with long legs’ has 71.8cm in waist circumference, waist height 99.7cm, and hip circumference 93.6cm. And it has 77.7cm in hip height, waist to hip length 22.9cm, and knee height 44.3cm.
According to lower body size on cluster, ‘Obesity’ is bigger than ‘Small hip’ and ‘Slim with long legs’ as 77.7cm in waist circumference. Also, ‘Obesity’ has 97.4cm hip circumference, the largest among the clusters and ‘Small hip’ has 91.6cm hip circumference, the smallest. In abdomen circumference, ‘Obesity’ has 86.7cm, the biggest and ‘Slim with long legs’ has 76.4cm, the smallest, so they have about 10cm differences. In height, waist, hip, abdomen, knee, and crotch are all highest in ‘Slim and Long legs’, but ‘Obesity’ and ‘Small hip’ are almost same. In width, waist width of ‘Obesity’ is the widest and that of ‘Small hip’ and ‘Slim with long legs’ is similar. ‘Small hip’ has the smallest hip width and ‘Obesity’ has the largest. Abdomen width is shown the smallest in ‘Slim with long legs’ as 27.7cm and the largest in ‘Obesity’ as 31.5cm. In length, ‘Slim with long legs’ has the longest hip length as 22.9cm and ‘Obesity’ and ‘Small hip’ are alike. ‘Slim with long legs’ has much longer in outside and inside of leg length. ‘Slim with long legs’ is the tallest as 164.5cm, ‘Obesity’ is next as 160.9cm, and ‘Small hip’ is 159.4cm. In case of weight, ‘Small hip’ is the lightest as 49.6kg, ‘Slim with long legs’ is next as 52.1kg, and ‘Obesity’ is 55.4kg.
V.Discussion and Conclusions
This research has attempted to analyze the shape of the lower half of women in their early 20s in Shanghai, China using somatometry with a 3D full body scanner. The results offer a wealth of information to manufacturers who export clothes to China. Kong et al. (2011) analyzed six groups for the ratio of thickness and width of characteristic circumference of the body. Choi (2008) analyzed the height of the lower half, obesity of the lower half, circumference of knee to ankle, shape of hip, length of waist to crotch, shape of abdomen and circumference of thigh of four groups. However, this study analyzed factors relating to whole body size and classified three clusters. A factor analysis for the shape of the lower body found five distinct factors: obesity of the lower half, height of the lower half, height from waist to crotch, shape of the abdomen, and leg bone length. Based on these factor analysis results, three clusters were determined. The first group is an obesity type, which exhibits high values of obesity and has the lowest value for lower-half height. The second group is a small hip type, with average values for most of the factors but the lowest value of hip circumference, hip width, abdomen height, thigh circumference, and weight. The third group has a slim body with long legs, exhibiting high values of lower-half height and length from waist to crotch. The circumferences of the ‘obesity’ type are the largest, and those of ‘small hip’ and ‘slim body with long legs’ types are similar, except for the hip circumference. The ‘small hip’ type has the smallest hip circumference. All of the height measurements for the ‘slim body with long legs’ type are the largest. Compared to Korean women in early 20s, Chinese early 20s women have longer legs, large difference in the size of circumference of waist and hips, and higher hip height, which is western body shape. Korean women have greatly grown lower body and large hips, and Chinese women have relatively flat lower body. So, considering these results, to design the pants for Chinese adult women, it needs to be set differently with Korean women in size of waist, hips, hip height, length of pants, and so on. However, it should be careful to generalize the results of this research since this research was progressed targeting only female college students in Donghua University in Shanghai, China, and anthropometry was in 2008. Therefore, further studies for various ages and regions in China need to be followed.
