Digital Human Modeling


Human Body Shape Modeling and Analysis

Characterizing and understanding the human shape variation is essential for many applications. Examples include product design, clothing, and health monitoring. Traditional anthropometry uses simple tools, like calipers and tape measures, to measure linear distances between anatomical landmarks. These measurements do not provide a unique description of the body shape.  They are also subject to human errors. With the development of 3D laser scanning technology, it is now possible to capture the entire body surface. Thus, a new field, 3D Anthropometry, has emerged. The first 3D anthropometric survey is the CAESAR (Civilian American and European Surface Anthropometry Resource) project. About 6000 civilians in the US, Italy, and the Netherlands aged between 18 and 65 were scanned. 

The goal of this project is to analyze the variability of the human body shape. We performed statistical shape analysis on the CAESAR database. The main modes of variation were extracted from a small number of principle components. Some of them correspond to our intuitive body characteristics such as height, weight, muscularity, etc. Analyzing and visualizing these variations form the foundation of many applications such as designing of clothing, helmets, masks, and other products that fit the human body better. It may also help build shape priors that can be used for reconstructing, recognizing, tracking, and animating human shapes.

Shape variations induced by some of the main components


Related publications

Pengcheng Xi, WonSook Lee, and Chang Shu, “Analysis of Segmented Human Body Scans,” Graphics Interface 2007. [pdf]

Pengcheng Xi, WonSook Lee, and Chang Shu, “A Data-driven Approach to Human Body Cloning Using a Segmented Body Database,” Pacific Graphics 2007 (to appear). [pdf]

Pengcheng Xi, Chang Shu, and Marc Rioux, “Principal Components Analysis on 3-D Scanned Human Heads,” SIGGRAPH poster, 2007. [pdf]

Zouhour Ben Azouz, Prosenjit Bose, Chang Shu, Stefanie Wuhrer, “Approximations of Geodesic Distances for Incomplete Triangular Manifolds,”
In Proceedings of the Canadian Conference on Computational Geometry (CCCG), 2007.
[pdf] [tech report] [conference talk]

T. Asano, P. Bose, P. Carmi, A. Maheshwari, Chang Shu, M. Smid, S. Wuhrer.
Linear-Space Algorithms for Distance Preserving Embedding.
In Proceedings of the Canadian Conference on Computational Geometry (CCCG), 2007.
[pdf] [conference talk]

Zouhour Ben Azouz, Chang Shu, Anja Mantel, "Automatic Locating of Anthropometric Landmarks on 3D Human Models," Third International Symposium on 3D Data Processing, Visualization and Transmission (3DPVT 2006). Chapel Hill, North Carolina, USA. June 13-16, 2006. [pdf]

Zouhour Ben Azouz, Marc Rioux, Chang Shu, and Richard Lepage, "Characterizing Human Shape Variation Using 3-D Anthropometric Data," The Visual Computer – International Journal of Computer Graphics, volume 22, number 5. 2005. pp. 302-314. [pdf]