摘要
Three-dimensional （3-D） surface anthropometry can provide much more useful information for many applications such as ergonomic product design than traditional individual body dimension measurements. However, the traditional definition of the percentile calculation is designed only for 1-D anthropometric data estimates. The same approach cannot be applied directly to 3-D anthropometric statistics otherwise it could lead to misinterpretations. In this paper, the influence of alignment references on 3-D anthropometric statistics is analyzed mathematically, which shows that different alignment reference points （for example, landmarks） for translation alignment could result in different object shapes if 3-D anthropometric data are processed for percentile values based on coordinates and that dimension percentile calculations based on coordinate statistics are incompatible with those traditionally based on individual dimensions.
Three-dimensional （3-D） surface anthropometry can provide much more useful information for many applications such as ergonomic product design than traditional individual body dimension measurements. However, the traditional definition of the percentile calculation is designed only for 1-D anthropometric data estimates. The same approach cannot be applied directly to 3-D anthropometric statistics otherwise it could lead to misinterpretations. In this paper, the influence of alignment references on 3-D anthropometric statistics is analyzed mathematically, which shows that different alignment reference points （for example, landmarks） for translation alignment could result in different object shapes if 3-D anthropometric data are processed for percentile values based on coordinates and that dimension percentile calculations based on coordinate statistics are incompatible with those traditionally based on individual dimensions.
基金
Supported by the Liberty Mutual Research Institute for Safety, USA (No. CSR03-03)
