Variational Motion Segmentation with Level Sets
We suggest a variational method for the joint estimation of optic flow and the segmentation of the image into regions of similar motion. It makes use of the level set framework following the idea of motion competition, which is extended to non-parametric motion. Moreover, we automatically determine an appropriate initialization and the number of regions by means of recursive two-phase splits with higher order region models. The method is further extended to the spatiotemporal setting and the use of additional cues like the gray value or color for the segmentation. It need not fear a quantitative comparison to pure optic flow estimation techniques: For the popular Yosemite sequence with clouds we obtain the currently most accurate result. We further uncover a mistake in the ground truth. Coarsely correcting this, we get an average angular error below 1 degree.
KeywordsComputer Vision Ground Truth Motion Estimation Active Contour Coarse Scale
Unable to display preview. Download preview PDF.
- 2.Amiaz, T., Kiryati, N.: Dense discontinuous optical flow via contour-based segmentation. In: Proc. International Conference on Image Processing, Genoa, Italy, September 2005, vol. 3, pp. 1264–1267 (2005)Google Scholar
- 3.Amiaz, T., Kiryati, N.: Piecewise-smooth dense optical flow via level sets. Technical Report VIA-2005-6-2, Vision and Image Analysis Laboratory, School of Electrical Engineering, Tel Aviv University, Israel (June 2005)Google Scholar
- 9.Bruhn, A., Weickert, J.: Towards ultimate motion estimation: Combining highest accuracy with real-time performance. In: Proc. 10th International Conference on Computer Vision, October 2005, pp. 749–755. IEEE Computer Society Press, Beijing (2005)Google Scholar
- 18.Mémin, E., Pérez, P.: A multigrid approach for hierarchical motion estimation. In: Proc. 6th International Conference on Computer Vision, Bombay, India, pp. 933–938 (1998)Google Scholar
- 20.Mumford, D., Shah, J.: Boundary detection by minimizing functionals, I. In: Proc. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Francisco, CA, June 1985, pp. 22–26. IEEE Computer Society Press, Los Alamitos (1985)Google Scholar
- 23.Papenberg, N., Bruhn, A., Brox, T., Didas, S., Weickert, J.: Highly accurate optic flow computation with theoretically justified warping. Technical Report 124, Dept. of Mathematics, Saarland University, Saarbrücken, Germany (January 2005); International Journal of Computer Vision (to appear)Google Scholar