CVFX Lecture 13: Optical flow

ECSE-6969 Computer Vision for Visual Effects
Rich Radke, Rensselaer Polytechnic Institute
Lecture 13: Optical flow (3/6/14)

0:00:02 Optical flow
0:00:58 Motion vectors
0:01:56 The brightness constancy assumption
0:05:44 The Horn-Schunck method
0:18:07 Hierarchical Horn-Schunck
0:40:25 The Lucas-Kanade method
0:45:54 Refinements and extensions
0:50:18 Smoothness along edges
0:53:52 Robust cost functions
0:58:12 Cross-checking
1:02:35 Layered flow
1:05:31 Large-displacement optical flow
1:08:13 Human-assisted motion annotation
1:11:31 Optical flow benchmarking
1:13:16 Optical flow for visual effects

Follows Section 5.3 of the textbook. http://cvfxbook.com

Note: despite the class discussion about hierarchical optical flow, the algorithm as presented on p. 159 of the book is correct.

Key references:

T. Brox, A. Bruhn, N. Papenberg, and J. Weickert. High accuracy optical flow estimation based on a theory for warping. In European Conference on Computer Vision (ECCV), 2004.
http://dx.doi.org/10.1007/978-3-540-2...

A. Bruhn, J. Weickert, and C. Schnörr. Lucas/Kanade meets Horn/Schunck: Combining local and global optic flow methods. International Journal of Computer Vision, 61(3):211--31, Feb. 2005.
http://dx.doi.org/10.1023/B:VISI.0000...

D. Sun, S. Roth, and M. Black. Secrets of optical flow estimation and their principles. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), 2010.
http://dx.doi.org/10.1109/CVPR.2010.5...

C. Liu, W. Freeman, E. Adelson, and Y. Weiss. Human-assisted motion annotation. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), 2008.
http://dx.doi.org/10.1109/CVPR.2008.4...