3D 스캔 데이터 기반의 통계적 인체형상 모델링

Abstract

In this paper, we present a method for parametric human body shape modeling based on a statistical approach to generate realistic 3D human body shape from key anthropometric measurements such as height and weight. This method is composed of three steps. First, we construct a human body shape database for statistical analysis from a 3D body scan data of 120 male with a wide variety of body type. Second, we accomplish feature analysis of human body shape through the results from the statistical analysis using principal component analysis (PCA). Finally, we derive the correlation between the human body shapes and anthropometric measurements to synthesize a new human body shape from the database. The proposed method allows user to generate realistic 3D human body shapes according to user-specified key anthropometric parameters such as height and weight. Also, our method is possible the detail human body shape modeling with the control of several anthropometric measurements presented in this paper except key anthropometric parameters. And, the generated model is automatically segmented to 16 regions, which can be utilized in a variety of areas such as ergonomic design or user-tailored design.