Dynamic Anthropometry – De ! ning Protocols for Automatic Body Measurement

The paper presents the research on possibilities of protocol development for automatic computer-based determination of measurements on a 3D body model in defi ned dynamic positions. Initially, two dynamic body positions were defi ned for the research on dimensional changes of targeted body lengths and surface segments during body movement from basic static position into a selected dynamic body position. The assumption was that during body movement, specifi c length and surface dimensions would change signifi cantly from the aspect of clothing construction and functionality of a garment model. 3D body scanning of a female test sample was performed in basic static and two defi ned dynamic positions. 3D body models were processed and measurement points were defi ned as a starting point for the determination of characteristic body measurements. The protocol for automatic computer measurement was defi ned for every dynamic body position by the systematic set of activities based on determined measurement points. The verifi cation of developed protocols was performed by automatic determination of defi ned measurements on the test sample and by comparing the results with the conventional manual measurement.


Introduction
Anthropometric human body measurement is the basis for obtaining accurate data on human body measures which are necessary for the construction of clothing block patterns [1]. Body measurements using conventional measuring instruments have been increasingly replaced by the use of modern computer technologies, such as various types of 3D body scanners [2,3]. In addition to the determination of linear body measures, as the most commonly used data in the garment industry and on which the conventional garment construction is based, 3D scans are used to obtain body shape data, anthropometric relationships of individual body parts, deviations from normal proportions, and body posture [4,5].
is enables the relevant data, necessary for the computer construction and the adjustment of garment patterns according to the individual anthropometric body characteristics of a particular subject [6,7]. 3D scanning technology can also be used for obtaining data on dynamic anthropometry, which is especially important when developing clothing for special purposes with high demands on functionality and t [8]. Depending on the body position and motion, body surface areas are deforming and body measurements are changing, which demands additional ease allowances on correspondent segments of the garment [9,10]. e international standard ISO 20685 has been developed to ensure the comparability of body measurements that are determined by standard ISO 7250 (Basic human body measurements of technological design) and ISO 8559 (Garment construction and anthropometric survey-body dimensions), obtained by using di erent types of 3D body scanners [11]. However, the accompanying computer programs applied for measurement on the obtained 3D point clouds at the end of the scanning process at this point enable automatic measurement procedure only in the basic standing body position. If it is necessary to determine body measurements in any of the dynamic positions, it is performed interactively, with precision measurement being largely dependent on the user/measurer. Such a measurement method is not appropriate for serial measurements of test subjects since the positioning of the measurement points has to be repeated for each subject in the same way in order to determine measurements. During manual interactive positioning it is very di cult to achieve the measurement precision and comparability of the measurement results for a sample of test subjects. In that sense, the paper presents the method of de ning the protocols for body measurement in two dynamic body positions as a starting point for the improvement of the clothing construction methods as well as the design and development of functional garment models.

Defi ning dynamic positions and 3D body scaning of female test subjects
Since the main objective of the research was to perform the analysis of the targeted lengths dimensions and body surface segments in the static and dynamic positions, two dynamic body positions were initially selected for 3D body scanning of female test subjects, together with standard static position, Figure 1.

3D models computer processing
Every particular 3D point cloud, i.e. body model of every test subject in static and dynamic positions was computer processed using the so ware Anthroscan 2016 (3.4.0.). Processing included closing body models surfaces and creating one-layered closed polygonal model.   Table 1, and 22 measurements for position 2, Table 2.  Sitting height, Figure 2e 2.

Defi ning of measurement points and characteristic body dimensions
Eye height 3.

Defi ning protocol for automatic measurement of body in dynamic position
When de ning a protocol for automatic measurement, it is primarily necessary to load each scanned 3D body model in dynamic position into the same le with the appropriate 3D model in the static position, Figure 3. A created combined scans le needs to be stored as a protocol con guration within the Anthroscan program.

Figure 3: Combined scans data le
In this way, the connection between the scanned models in the dynamic positions and the anthropometric points de ned by the ISO 7520 standard is achieved. Loading of the created le into the program enables de ning of target dimensions in the dynamic position, whereby the coordinates of anthropometric points between which the measurement will be performed are de ned directly on the 3D model. e positions of the de ned anthropometric points are stored as a le of measurements within the dened protocol. e protocol must be de ned separately for each selected dynamic position, a er which the automatic measurement of scanned body models in dynamic positions is enabled by selecting the desired protocol within the program menu. A combination of protocols can be achieved by proper naming of the scanned model les within the particular group with the protocol as a name pre x of the group. Based on the name pre x, the system will automatically apply a targeted protocol without the need to select it. In this case, it is necessary to create bmp image les of the targeted positions based on which the system will recognize which scan model within the combined scans le is appropriate for the application of each protocol. is also enables the application of multiple protocols in the same measurement procedure. Despite the fact that the protocols for measurement in dynamic positions are de ned separately based on group naming, the system will link proper scan model with the targeted protocol. e protocol also enables automatic creation of a measurement table, i.e. html le with measurements and coordinates of anthropometric points. It is also possible to create a speci c measurement menu only with the selection of the measurements that will be included in further analysis, Figure 4.    Using the de ned protocols, the coordinates of 14 characteristic measurement points on position 1, and 22 points coordinates of position 2 were obtained for every test subject and saved as a measurement .bmf le, Table 3 and Table 4. e measurement le enabled automatic measurement procedure between de ned points according to the measurement list or in this case the customized measurement menu. For this purpose, measurements selection was performed and a new measurements menu was created for analysis of the hip widths values between posture 1 and 2, Figure 4. Beside the hip width measurements in dynamic postures de ned by the new protocols, the menu included basic body circumferences according to Standard. at way, testing of possibilities for combined measurements in both static and dynamic conditions was performed. Test subjects were grouped according to garment size, which was based on the breast girth measurement, to test the correlation between hip width changes and body size. e determined changes in hip widths are in the range from the 0.7 to 6.0 cm, Table 5. Legend: GS -garment size, TS -test subject, BG -breast girth, HG -hip girth, HW1 -hip width on position 1, HW2 -hip width on position 2, ∆HW -di erence in hip widths measurements on position 1 and position 2, r -correlation coe cient e correlation analysis of the determined changes in hip width measurement showed very low correlation with breast girth, i.e. garment size and no correlation with hip girth, Figure 5 and Figure 6.
is implies that hip measurement changes on different dynamic body positions cannot be predicted and classi ed using body size and body girths measurements, but the body morphology must be included in the analysis. t. Dimensional body changes must be taken into consideration while designing and constructing a garment with high criteria for functionality such as protective and sports garments. A dimensional change of body in dynamic positions is a complex issue that cannot be considered only from the aspect of basic anthropometric measurement, but the body morphology parameters must be taken into consideration. Creating automatic measurement protocols for dynamic positions provides greater precision in de ning positions of anthropometric points compared to manual processing of individual models. It reduces a measurer's mistake and is repeatable since the points are always positioned in the same way according to the measurement le de ned within the protocol. is enables comparison of the obtained measurement results since the measurements are always taken in the same manner, providing the possibilities for complex studies of body dimensional changes in dynamic conditions.