GPU-accelerated Real-time Markerless Human Motion Capture

Christian Rau, Guido Brunnett

2013

Abstract

We present a system for capturing human motions based on video data from multiple cameras. It realizes a 3-dimensional voxel-based reconstruction of the human together with an estimation of the pose of his complete body in each frame. The use of an underlying kinematic skeleton together with an idealized geometric model can guarantee a valid pose even in the face of occlusions caused by the incomplete spatial information gained from the cameras. The data-parallel nature of the used algorithms makes them well-suited for the implementation on modern graphics hardware. In this way the motion can be captured in real-time on a single PC despite the computation of a reconstruction accurate enough for a high quality pose estimation.

References

  1. Caillette, F. and Howard, T. (2004). Real-time markerless human body tracking with multi-view 3-d voxel reconstruction. In Proceedings of BMVC, pages 597- 606.
  2. Cheung, K.-M., Kanade, T., Bouguet, J.-Y., and Holler, M. (2000). A real time system for robust 3d voxel reconstruction of human motions. In Proceedings of the 2000 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 7800), pages 714-720.
  3. Corazza, S., Mündermann, L., Gambaretto, E., Ferrigno, G., and Andriacchi, T. P. (2010). Markerless motion capture through visual hull, articulated icp and subject specific model generation. International Journal of Computer Vision, 87(1):156-169.
  4. Hasenfratz, J.-M., Lapierre, M., Gascuel, J.-D., and Boyer, E. (2003). Real-time capture, reconstruction and insertion into virtual world of human actors. In Vision, Video and Graphics, pages 49-56. Elsevier.
  5. Kehl, R. and Gool, L. V. (2006). Markerless tracking of complex human motions from multiple views. Computer Vision and Image Understanding, 104(2):190- 209.
  6. Luck, J., Small, D., and Little, C. (2001). Real-time tracking of articulated human models using a 3d shapefrom-silhouette method. In Robot Vision, Lecture Notes in Computer Science, pages 19-26. Springer Verlag.
  7. Matusik, W., Buehler, C., and McMillan, L. (2001). Polyhedral visual hulls for real-time rendering. In Proceedings of the 12th Eurographics Workshop on Rendering Techniques, pages 115-126. Springer Verlag.
  8. Moeslund, T. B., Hilton, A., and Krüger, V. (2006). A survey of advances in vision-based human motion capture and analysis. Computer Vision and Image Understanding, 104(2):90-126.
  9. Satish, N., Harris, M., and Garland, M. (2009). Designing efficient sorting algorithms for manycore gpus. In Proceedings of the 2009 IEEE International Symposium on Parallel and Distributed Processing, pages 1-10. IEEE Computer Society.
  10. Sengupta, S., Harris, M., and Garland, M. (2008). Efficient parallel scan algorithms for gpus. Technical report, NVIDIA.
  11. Toyama, K., Krumm, J., Brummit, B., and Meyers, B. (1999). Wallflower: Principles and practice of background maintenance. In Proceedings of the Seventh IEEE International Conference on Computer Vision, pages 255-261.
Download


Paper Citation


in Harvard Style

Rau C. and Brunnett G. (2013). GPU-accelerated Real-time Markerless Human Motion Capture . In Proceedings of the International Conference on Computer Graphics Theory and Applications and International Conference on Information Visualization Theory and Applications - Volume 1: GRAPP, (VISIGRAPP 2013) ISBN 978-989-8565-46-4, pages 397-401. DOI: 10.5220/0004282203970401


in Bibtex Style

@conference{grapp13,
author={Christian Rau and Guido Brunnett},
title={GPU-accelerated Real-time Markerless Human Motion Capture},
booktitle={Proceedings of the International Conference on Computer Graphics Theory and Applications and International Conference on Information Visualization Theory and Applications - Volume 1: GRAPP, (VISIGRAPP 2013)},
year={2013},
pages={397-401},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0004282203970401},
isbn={978-989-8565-46-4},
}


in EndNote Style

TY - CONF
JO - Proceedings of the International Conference on Computer Graphics Theory and Applications and International Conference on Information Visualization Theory and Applications - Volume 1: GRAPP, (VISIGRAPP 2013)
TI - GPU-accelerated Real-time Markerless Human Motion Capture
SN - 978-989-8565-46-4
AU - Rau C.
AU - Brunnett G.
PY - 2013
SP - 397
EP - 401
DO - 10.5220/0004282203970401