3D Scanning of Dutch Military: Secular Trends in PCA for 18,000 Soldiers
conference paper
The Netherlands Military Forces started 3D whole body scanning of their recruits in the year 2003.
1D-body dimensions are derived from the scans to supply the best fitting size without human involvement. In order to investigate secular trends in body dimensions, the dataset of 3D scans was cleaned and analyzed for inter-individual differences using principle component (PC) analysis.
Changes in body dimensions and body posture were quantified over the period 2003-2012 and digital design extremes were constructed.
The original dataset of 18221 scans was cleaned to 15231 male and 2390 female scans (97%). A male and female model was constructed from 200 selected men and 200 selected women. Since the arm position was not standardized during scanning, it was necessary to separate the arms from the torso.
The analysis showed that men and women differed most for stature (PC1), mass and inner leg length (PC2) and mass without inner leg length (PC3). These PC values hardly changed over the years.
However, the PCs of the arms changed over the years due to changing preferences in posture.
The developed model enables quantification of the body shape and posture with only a small number of PCs. This means that the data can be stored and processed effectively. The generated design extremes may serve to optimize products related to the physical shape of soldiers.
1D-body dimensions are derived from the scans to supply the best fitting size without human involvement. In order to investigate secular trends in body dimensions, the dataset of 3D scans was cleaned and analyzed for inter-individual differences using principle component (PC) analysis.
Changes in body dimensions and body posture were quantified over the period 2003-2012 and digital design extremes were constructed.
The original dataset of 18221 scans was cleaned to 15231 male and 2390 female scans (97%). A male and female model was constructed from 200 selected men and 200 selected women. Since the arm position was not standardized during scanning, it was necessary to separate the arms from the torso.
The analysis showed that men and women differed most for stature (PC1), mass and inner leg length (PC2) and mass without inner leg length (PC3). These PC values hardly changed over the years.
However, the PCs of the arms changed over the years due to changing preferences in posture.
The developed model enables quantification of the body shape and posture with only a small number of PCs. This means that the data can be stored and processed effectively. The generated design extremes may serve to optimize products related to the physical shape of soldiers.
TNO Identifier
484684
Source title
4th International Conference and Exhibition on 3D Body Scanning Technologies, 3DBST 2013, 19 - 20 November 2013, Long Beach CA, USA