Automatic Generation of Statistical Shape Models in Motion

Conference Paper (2019)
Author(s)

Femke Danckaers (Universiteit Antwerpen)

Sofia Scataglini (Royal Military Academy, Military Hospital Queen Astrid)

Robby Haelterman (Royal Military Academy)

Damien Van Tiggelen (Military Hospital Queen Astrid)

T. Huysmans (Universiteit Antwerpen, TU Delft - Human Factors)

Jan Sijbers (Universiteit Antwerpen)

Research Group
Human Factors
DOI related publication
https://doi.org/10.1007/978-3-319-94223-0_16
More Info
expand_more
Publication Year
2019
Language
English
Research Group
Human Factors
Pages (from-to)
170-178
ISBN (print)
978-3-319-94222-3
ISBN (electronic)
978-3-319-94223-0

Abstract

Statistical body shape modeling (SBSM) is a well-known technique to map out the variability of body shapes and is commonly used in 3D anthropometric analyses. In this paper, a new approach to integrate movement acquired by a motion capture system with a body shape is proposed. This was done by selecting landmarks on a body shape model, and predicting a body shape based on features. Then, a virtual skeleton was generated relative to those landmarks. This skeleton was parented to a body shape, allowing to modify its pose and to add pre-recorded motion to different body shapes in a realistic way.

No files available

Metadata only record. There are no files for this record.