Page 47 - Anthropometry, Apparel Sizing and Design

Anthropometric methods for apparel design: Body measurement devices and techniques 43

right) that includes 90% of all the measured data points. The standard deviation of r i is
also used to evaluate the error of the spherical form measurement.
ISO 20685-2 also establishes a protocol for evaluating the repeatability of mea-
sured landmark locations using an anthropomorphic dummy representing the size
and shape of a natural human. Repeatability rather than the accuracy of landmark loca-
tion is evaluated because shape of anthropomorphic dummies is too complicated to
give the actual values using a CMM. Marker stickers are placed on landmarks pre-
marked on the dummy. The dummy is scanned 10 times at slightly different positions.
Landmark coordinates are calculated for the 10scans. Landmark location data from
10scans are superimposed simultaneously. The error is calculated as the distance
between corresponding landmark locations for all possible pairs of landmark locations
data (N ¼ 10 C 2 ¼45). Mean error and the maximum error are reported. Examples of
evaluation results using a ball with the diameter of 120.01593mm and an anthropo-
morphic dummy are published in Kouchi et al. (2012).
Occluded areas due to the complicated body shape and quality of scan-derived
landmark locations are issues specific to anthropometry. Occluded areas cannot be
evaluated using a test object. ISO 20685-2 provides a protocol for evaluating the
occluded area using human subjects.
It may not be easy to access an artifact calibrated using a CMM. A hard and non-
deformable ball of appropriate surface treatment and size can be used for daily ver-
ification. This can be done by measuring the test object using a 3-D body scanner to be
evaluated and comparing the measured diameter with the diameter measured with a
calibrated vernier caliper.
The following information should be considered when selecting a 3-D body scan-
ner: scan volume, time necessary for one scan, scan direction, resolution, accuracy,
function to capture the texture, function to capture premarked landmark locations
and its method (manual or automatic), number of cameras arranged in different direc-
tions, function to merge data obtained by different cameras, function to calculate 1-D
measurements, and function to generate a homologous model. The final decision
depends on the required type(s) of data (1-D measurements, landmark location,
and surface shape), accuracy of data, and funds (time and space allowed to obtain
required data, price of a 3-D body scanner, and running cost).

2.4 International standards related to anthropometric
methods

It is implicitly assumed that measurements from different surveys are taken
usingexactly thesamemethodwhenthe measurement name is identical. Unfortu-
nately, this is not always true. Measurement items named identically but defined
differently or named differently but defined identically cause confusion. To avoid
such unnecessary confusion, textbooks and standards are used as references. The
most frequently used reference is an anthropometry textbook, Martin’s textbook
of anthropology (Martin and Knußmann, 1988), but this textbook does not focus
on anthropometry for garment design. There are several international standards