Escolar Documentos
Profissional Documentos
Cultura Documentos
Functional clothing encompasses a wide range of apparels such as protective equipments, functional garments and fireretardant clothing. Depending on the application, they must adapt to the shape of the human body, should not interfere with
the body motion and should isolate the body from a potentially hazardous environment. In order to achieve these objectives,
it is imperative to characterize a body or a body part with the most representative measurements and to understand their
interrelationships. More recently, it has been possible to acquire the three-dimensional shape of the human body with full
body scanners. Such a three-dimensional shape is of course highly desirable for functional clothing, because it contains
detailed information about the whole shape; for instance, a protective mask is most likely to be efficient if the shape around
the mouth is known precisely. Consequently, it is essential to be able to analyze and compare these shapes and to understand
their relations with traditional anthropometric measurements which are still widely used by the apparel industry. This paper
describes our results when analyzing an anthropometric database, using both anthropometric measures and 3D data. Our
results indicate that these two sets of data offer complimentary perspectives of the human body and it's measurements, in
terms of the resultant sets of clusters and associated archetypes. Through combining these complimentary perspectives, the
goal of designing well-fitting functional apparel may be realized.
Keywords: Anthropometry, Archetype, Clustering, Functional clothing, Three-dimensional shape
1 Introduction
The design of functional clothing that fits the subject
well is of crucial importance in many domains, ranging
from protective clothing for hazardous waste disposal,
pilot helmet design to astronaut suit design. In such
domains, the apparel should be comfortable, and
should not hinder the task at hand. These clothing
should be based on real anthropometric data, and also
incorporate information regarding the shape of the
human body, in order to ensure a close fit.
The CAESARTM database is an anthropometric
database that contains detailed data regarding human
subjects that reside in the USA, Italy and the
Netherlands. The database contains numerous
anthropometric measures and demographic profiles. In
addition, it contains the 3D body scans of all subjects.
This database thus presents us with the opportunity to
profile the typical groupings of human bodies, based on
the anthropometric data as well as 3D data. This paper
presents the results of our analysis, when aiming to find
the archetypes of typical groupings or clusters, for
functional clothing design.
_____________
a
To whom all the correspondence should be addressed.
E-mail: eric.paquet@nrc-cnrc.gc.ca
337
338
XX-Large
93.1 (6.0)
Male
96.5 (4.8)
82.5 (6.9)
86.5 (5.7)
96.3 (7.4)
97.4 (5.7)
112.5 (9.1)
95.3 (4.6)
45.4 (2.4)
98.5 (3.6)
47.4 (2.5)
103.1 (4.5)
50.3 (2.7)
108.1 (4.0)
51.0 (2.3)
116.3 (6.2)
55.2 (3.4)
60.3 (3.3)
65.4 (2.6)
62.4 (3.0)
67.2 (3.2)
65.6 (3.6)
173.0 (6.6)
44.0 (1.6)
67.99 (6.7)
126 (22 %)
184.8 (4.6)
46.6 (1.8)
77.79 (5.53)
173 (31 %)
176.1 (5.6)
47.8 (2.0)
90.2 (7.93)
139 (25 %)
193.2 (6.2)
48.8 (2.0)
97.25(7.67)
82 (14 %)
188.1 (8.3)
52.0 (3.4)
119.79 (13.24)
47 (8 %)
90.7 (5.9)
74.1 (6.4)
92.1 (5.0)
76.4 (5.9)
Female
98.4 (6.2)
83.5 (7.1)
106.1 (6.7)
90.7 (8.0)
117.5 (8.3)
103.3 (9.4)
120.6 (9.2)
107.9 (11.0)
97.2 (5.1)
55.1 (2.4)
160.0 (4.7)
40.3 (1.7)
57.38(5.94)
101.2 (5.4)
58.5 (1.8)
169.7 (3.7)
41.7 (1.7)
64 (5.62)
106.4 (5.1)
62.0 (2.2)
176.9 (4.8)
44.0 (2.1)
74.97(6.58)
110.1 (5.8)
56.8 (1.9)
162.4 (5.1)
43.6 (2.4)
77.61(6.03)
117.5 (7.7)
58.6 (2.2)
167.1 (4.3)
47.0 (2.9)
95.25 (9.89)
121.6 (10.4)
63 (2.0)
176.9 (5.6)
47.6 (2.9)
106.27 (13.11)
125 (18 %)
83 (12 %)
39 (6 %)
Number of subjects
130 (19 %)
198 (28 %)
125 (18 %)
Mean values are given in cm along with the standard deviation in parentheses.
107.8 (6.7)
104.9 (6.2)
121.2 (7.0)
Fig. 1Cluster centroids for Dutch male population [ (a) Small, (b) Medium, (c) Large, (d) X-Large and (e) XX-Large]
339
Fig. 2 Cluster centroids for Dutch female population [ (a) X-Small, (b) Small, (c) Medium, (d) Large, (e) X-Large and (f) XX-Large]
Fig. 3Cluster visualization of 3D data for the Dutch population (a) clustering of the male population and (b) clustering of the female
population
340
Number of subjects
Male
Female
135 (25%)
79 (15%)
166 (31%)
106 (20%)
56 (10%)
132 (20%)
63 (10%)
103 (16%)
149 (22%)
147 (22%)
69 (10%)
Fig. 4Cluster centroids from 3D data for Dutch males [ (a) Small,
(b) Medium, (c) Large, (d) X-Large and (e) XX-Large]
Fig. 5Cluster centroids from 3D data for Dutch females [ (a) X-Small, (b) Small,(c ) Medium, (d) Large, (e) X-Large and (f) XX-Large]
341
Chest circumference
Waist circumference
Hip circumference
Neck base circumference
Arm shoulder to wrist
Stature
Shoulder breadth
Weight
96.6 cm
88.5 cm
97.5 cm
49.0 cm
61.3 cm
169.2 cm
43.6 cm
73.21 kg
Fig. 7 Shape comparison with the Small centroid. The subject that
decreases three sizes ( a), and the Small centroid from 3D data ( b)
Value
Value
103.0 cm
94.1 cm
101.7 cm
49.5 cm
66.9 cm
195.6 cm
47.7 cm
89.99 kg
342
Fig. 8 Shape comparison with the Small centroid. The subject that
decreases four sizes (a), and the Small centroid from 3D data (b)
Fig. 9 Shape comparison with the X-Large centroid. The subject that
increases four sizes ( a), and the X-Large centroid from 3D data (b)
Value
104.6 cm
87.8 cm
98.4 cm
122.8 cm
65.4 cm
194.8 cm
45.1 cm
103.42 kg
Value
Bust circumference
Under bust circumference
Waist circumference
Hip circumference
Shoulder to wrist
Stature
Shoulder breadth
Weight
94.2 cm
72.5 cm
75.4 cm
102.7 cm
52.7 cm
152.0 cm
42.9 cm
62.18 kg
taken into account, as observed from the 3D shapebased cluster membership. That is, within a specific
anthropometric cluster, the various shapes, as
obtained from the 3D shape-based results, may be
used to create a set of variations taking the common
body shapes (e.g. a large person with narrow back
and/or sloped shoulder shapes) into account.
6 Conclusion
The design of well-fitting functional clothing is
crucial to ensure the safety of personnel who works
in
dangerous
and
potentially
hazardous
environments. The study indicates that the
anthropometric and 3D data yields complimentary
results, which should be used together in order to
find the best fit. In particular, the 3D shape-based
clusters should be used to refine and complement the
anthropometric archetypes, in order to model typical
body shapes. Future work will include the testing of
our results in a real world setting. We are also
interested to extend this work to study the
anthropometry of the disabled. For instance, the
design of artificial limbs and clothing to facilitate
wheelchair navigation comes to mind.
References
1 Tan P-N, Steinbach M & Kumar V, Introduction to Data
Mining (Addison Wesley, Chichester, England), 2005, 1.
343