Note: Your custom schedule will not be saved unless you create a new account or login to an existing account.
| Paper: | IMDSP-P5.7 |
| Session: | Motion Estimation |
| Time: | Wednesday, May 19, 09:30 - 11:30 |
| Presentation: |
Poster |
| Topic: |
Image and Multidimensional Signal Processing: M-D Signal Processing Theory and Methods |
| Title: |
EMPIRICAL CHOICE OF SMOOTHING PARAMETERS IN ROBUST OPTICAL FLOW ESTIMATION |
| Authors: |
Mingren Shi; University of New South wales | | |
| | Victor Solo; University of New South Wales | | |
| Abstract: |
Optical flow estimation algorithms such as theLukas-Kanade method and Horn and Schunk methodrequire selection of a tuning parameter. In the former case a neighbourhoodsize, in the latter, a penalty parameter. Selectionof these tuning parameters is difficult in generalbut has a profound effect on the results. So automatic methods of selection are of great interest.In previous work we have developed selection methods for the algorithms above. Here we develop a selection procedurefor a robust version of the Lukas-Kanade method. This isa non-trivial task since the robust algorithm is nonlinear. |
| |
| Back | |