RECOGNITION OF DYNAMIC VIDEO CONTENTS BASED ON MOTION TEXTURE STATISTICAL MODELS

Tomas Crivelli, Bruno Cernuschi-Frias, Patrick Bouthemy, Jian-Feng Yao

Abstract

The aim of this work is to model, learn and recognize, dynamic contents in video sequences, displayed mostly by natural scene elements, such as rivers, smoke, moving foliage, fire, etc. We adopt the mixed-state Markov random fields modeling recently introduced to represent the so-called motion textures. The approach consists in describing the spatial distribution of some motion measurements which exhibit values of two types: a discrete component related to the absence of motion and a continuous part for measurements different from zero. Based on this, we present a method for recognition and classification of real motion textures using the generative statistical models that can be learned for each motion texture class. Experiments on sequences from the DynTex dynamic texture database demonstrate the performance of this novel approach.

References

  1. Besag, J. (1974). Spatial interaction and the statistical analysis of lattice systems. Journal of the Royal Statistical Society. Series B, 36:192-236.
  2. Bouthemy, P., Hardouin, C., Piriou, G., and Yao, J. (2006). Mixed-state auto-models and motion texture modeling. Journal of Mathematical Imaging and Vision, 25:387-402.
  3. Cernuschi-Frias, B. (2007). Mixed-states markov random fields with symbolic labels and multidimensional real values. Rapport de Recherche INRIA.
  4. Crivelli, T., Cernuschi, B., Bouthemy, P., and Yao, J. (2006). Segmentation of motion textures using mixed-state markov random fields. In Proceedings of SPIE, volume 6315, 63150J.
  5. Doretto, G., Chiuso, A., Wu, Y., and Soatto, S. (2003). Dynamic textures. Intl. Journal of Comp. Vision, 51(2):91-109.
  6. Fablet, R. and Bouthemy, P. (2003). Motion recognition using non-parametric image motion models estimated from temporal and multiscale co-ocurrence statistics. IEEE Trans. on Pattern Analysis and Machine Intelligence, 25(12):1619-1624.
  7. Fazekas, S. and Chetverikov, D. (2005). Normal versus complete flow in dynamic texture recognition: a comparative study. In Texture 2005: 4th international workshop on texture analysis and synthesis. ICCV'05, Beijing, pages 37-42.
  8. Geman, S. and Geman, D. (1984). Stochastic relaxation, gibbs distributions, and the bayesian restoration of images. IEEE Transactions on Pattern Analysis and Machine Intelligence, 6:721-741.
  9. Lu, Z., Xie, W., Pei, J., and Huang, J. (2005). Dynamic texture recognition by spatio-temporal multiresolution histograms. In IEEE Workshop on Motion and Video Computing (WACV/MOTION), pages 241-246.
  10. Peteri, R. and Chetverikov, D. (2005). Dynamic texture recognition using normal flow and texture regularity. In Proc. of IbPRIA, Estoril, pages 223-230.
  11. Peteri, R., Huiskes, M., and Fazekas, S. (2005). Dyntex: A comprehensive database of dynamic textures. http://www.cwi.nl/projects/dyntex/index.html.
  12. Saisan, P., Doretto, G., Wu, Y., and Soatto, S. (2001). Dynamic texture recognition. In Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition. CVPR'01, Hawaii, pages 58-63.
  13. Salzenstein, F. and Pieczynski, W. (1997). Parameter estimation in hidden fuzzy markov random fields and image segmentation. Graph. Models Image Process., 59(4):205-220.
  14. Vidal, R. and Ravichandran, A. (2005). Optical flow estimation and segmentation of multiple moving dynamic textures. In Proc. of CVPR'05, SanDiego, volume 2, pages 516-521.
  15. Yuan, L., Wen, F., Liu, C., and Shum, H. (2004). Synthesizing dynamic textures with closed-loop linear dynamic systems. In Proc. of the 8th European Conf. on Computer Vision, ECCV'04, Prague, volume LNCS 3022, pages 603-616.
Download


Paper Citation


in Harvard Style

Crivelli T., Cernuschi-Frias B., Bouthemy P. and Yao J. (2008). RECOGNITION OF DYNAMIC VIDEO CONTENTS BASED ON MOTION TEXTURE STATISTICAL MODELS . In Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2008) ISBN 978-989-8111-21-0, pages 283-289. DOI: 10.5220/0001078102830289


in Bibtex Style

@conference{visapp08,
author={Tomas Crivelli and Bruno Cernuschi-Frias and Patrick Bouthemy and Jian-Feng Yao},
title={RECOGNITION OF DYNAMIC VIDEO CONTENTS BASED ON MOTION TEXTURE STATISTICAL MODELS},
booktitle={Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2008)},
year={2008},
pages={283-289},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001078102830289},
isbn={978-989-8111-21-0},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2008)
TI - RECOGNITION OF DYNAMIC VIDEO CONTENTS BASED ON MOTION TEXTURE STATISTICAL MODELS
SN - 978-989-8111-21-0
AU - Crivelli T.
AU - Cernuschi-Frias B.
AU - Bouthemy P.
AU - Yao J.
PY - 2008
SP - 283
EP - 289
DO - 10.5220/0001078102830289