Advertisement

Image and Video Segmentation by Anisotropic Kernel Mean Shift

  • Jue Wang
  • Bo Thiesson
  • Yingqing Xu
  • Michael Cohen
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3022)

Abstract

Mean shift is a nonparametric estimator of density which has been applied to image and video segmentation. Traditional mean shift based segmentation uses a radially symmetric kernel to estimate local density, which is not optimal in view of the often structured nature of image and more particularly video data. In this paper we present an anisotropic kernel mean shift in which the shape, scale, and orientation of the kernels adapt to the local structure of the image or video. We decompose the anisotropic kernel to provide handles for modifying the segmentation based on simple heuristics. Experimental results show that the anisotropic kernel mean shift outperforms the original mean shift on image and video segmentation in the following aspects: 1) it gets better results on general images and video in a smoothness sense; 2) the segmented results are more consistent with human visual saliency; 3) the algorithm is robust to initial parameters.

Keywords

Image Segmentation Video Data Shift Point Video Segmentation Color Domain 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

References

  1. 1.
    Comaniciu, D., Meer, P.: Mean shift analysis and applications. In: Proc. IEEE Int. Conf. on Computer Vision, Greece, pp. 1197–1203 (1999)Google Scholar
  2. 2.
    Comaniciu, D., Ramesh, V., Meer, P.: Real-time tracking of non-rigid objects using mean shift. In: Proc. IEEE Int. Conf. on Computer Vision and Pattern Recognition, pp. 142–151 (2000)Google Scholar
  3. 3.
    DeMenthon, D., Megret, R.: The variable bandwidth mean shift and data-driven scale selection. In: Proc. IEEE 8th Int. Conf. on Computer Vision, Canada, pp. 438–445 (2001)Google Scholar
  4. 4.
    Comaniciu, D.: An Algorithm for Data-Driven Bandwidth Selection. IEEE Transactions on Pattern Analysis and Machine Intelligence 25(2) (February 2003)Google Scholar
  5. 5.
    Comaniciu, D., Meer, P.: Mean shift: a robust approach toward feature space analysis. IEEE Trans. on PAMI, 603–619 (2002)Google Scholar
  6. 6.
    DeMenthon, D., Megret, R.: Spatio-temporal segmentation of video by hierarchical mean shift analysis. In: Proc. IEEE Int. Conf. on Computer Vision and Pattern Recognition, pp. 142–151 (2000)Google Scholar
  7. 7.
    Fukunaga, K., Hostetler, L.: The estimation of the gradient of a density function, with applications in pattern recognition. IEEE Trans. Information Theory 21, 32–40 (1975)zbMATHCrossRefMathSciNetGoogle Scholar
  8. 8.
    Lorensen, W.E., Cline, H.E.: Marching Cubes: a high resolution 3D surface reconstruction algorithm. In: Proc. ACM SIGGRAPH 1987, pp. 163–169 (1987)Google Scholar
  9. 9.
    Megret, R., DeMenthon, D.: A Survey of Spatio-Temporal Grouping Techniques. Technical report: LAMP-TR-094/CS-TR-4403, University of Maryland, College Park (1994)Google Scholar
  10. 10.
    Pal, N.R., Pal, S.K.: A review on image segmentation techniques. Pattern Recognition 26(9), 1277–1294 (1993)CrossRefGoogle Scholar
  11. 11.
    Skarbek, W., Koschan, A.: Colour Image Segmentation: A survey. Technical report, Technical University Berlin (1994)Google Scholar
  12. 12.
    Singh, M., Ahuja, N.: Regression Based Bandwidth Selection for Segmentation using Parzen Windows. In: Proc. IEEE International Conference on Computer Vision, vol. 1, pp. 2–9 (2003)Google Scholar
  13. 13.
    Wand, M., Jones, M.: Kernel Smoothing, p. 95. Chapman & Hall, Boca Raton (1995)zbMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • Jue Wang
    • 1
  • Bo Thiesson
    • 1
  • Yingqing Xu
    • 1
  • Michael Cohen
    • 1
  1. 1.Microsoft Research (Asia and Redmond) 

Personalised recommendations