Authors:

Ann-Malin SCHMIDT 1, Ingrid PERAZA 1,2, Yordan KYOSEV 1

1 Chair of Development and Assembly of Textile Products, ITM, TU Dresden, Germany;
2 Ghent University, Belgium

DOI:

https://doi.org/10.15221/24.40

Full paper:

PDF

Keywords:

4D scanning, SMPL, OSSO, Skeleton fitting

Abstract:

4D scan data capture the body's surface in motion, enabling analysis of body deformation, skin stretching, and clothing fit, but are limited to surface-level information. Previous work used 4D scan data to develop individualized Finite Element (FE) models for digital clothing fitting, but these models lack a bone structure and require labor-intensive manual placement or costly CT scans.
An automated method for fitting a skeleton to 4D scan data has been tested and evaluated on three subjects in two different poses. This method involves converting 4D scan data into SMPL models, followed by the automatic application of a skeleton using the open-source Python module OSSO.
The fitting process of the SMPL model demonstrated high accuracy and repeatability, with mesh differences of less than 5 mm for static A-Pose configurations. Although the SMPL model simplifies body surface details, resulting in a slightly slimmer appearance, it successfully supports accurate skeleton fitting to the original 4D scan data. In dynamic poses from 4D scans, inaccuracies and low repeatability were observed for the SMPL fitting. The OSSO-based method efficiently placed bone poses across all tested poses with only minor penetrations noted in areas such as the fingers, legs, and ribs. Integrating the OSSO skeleton with 4D scan data produced a model that effectively combined accurate body surface representation with a detailed skeleton. Nonetheless, pose discrepancies between the SMPL model and the 4D scan data resulted in some intersections for the skeleton mesh, which could lead to errors in Finite Element (FE) models. Minor adjustments, such as rotating joints, improved the fit of the skeleton. Overall, while the method shows promise, further refinements in SMPL model fitting process, especially for complex poses, are needed to enhance its potential for improving the precision of individualized FE models.

How to Cite (MLA):

Schmidt, Ann-Malin et al., "Evaluation of Automated Skeleton Fitting to 4D Human Body Scan Data Using Open-Source SMPL- and OSSO Models", 3DBODY.TECH Journal, vol. 1, Oct. 2024, #40, https://doi.org/10.15221/24.40.

Presentation:

VIDEO not available

Details:

Volume/Issue: 3DBODY.TECH Journal - Vol. 1, 2024
Paper: #40
Published: 2024/10/30
Presented at: 3DBODY.TECH 2024, 22-23 Oct. 2024, Lugano, Switzerland
Proceedings: 3DBODY.TECH 2024 Proceedings

License/Copyright notice

Copyright © 2024 by the author(s).
This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
The papers appearing in the journal reflect the author's opinions. Their inclusion in the volumes does not necessary constitute endorsement by the editor or by the publisher.

Note: click the + on the top left of the page to open/close the menu.