Parametric human body shape modeling framework for human-centered product design

The objective of this study is the development of a novel parametric human body shape modeling framework for integration into various product design applications. Our modeling framework is comprised of three phases of database construction, statistical analysis, and model generation. During the database construction phase, a 3D whole body scan data of 250 subjects are obtained, and their data structures are processed so as to be suitable for statistical analysis. Using those preprocessed scan data, the characteristics of the human body shape variation and their correlations with several items of body sizes are investigated in the statistical analysis phase. The correlations obtained from such analysis allow us to develop an interactive modeling interface, which takes the body sizes as inputs and returns a corresponding body shape model as an output. Using this interface, we develop a parametric human body shape modeling system and generate body shape models based on the input body sizes. In our experiment, our modeler produced reasonable results having not only a high level of accuracy but also fine visual fidelity. Compared to other parametric human modeling approaches, our method contributes to the related field by introducing a novel method for correlating body shape and body sizes and by establishing an improved parameter optimization technique for the model generation process.

Research highlights
► We propose methodology for modeling the 3D body shape from user-specified body sizes.
► Body shape database is constructed based on 3D whole body scan data.
► Modeling parameters are extracted from statistical analysis on various body shapes.
► Method to find the optimal shape parameters satisfying input body sizes is developed.