Lei Zhang¡¯s homepage
Scale and Orientation Adaptive
Mean Shift Tracking
by Jifeng Ning, Lei Zhang, David
Zhang and Chengke Wu
To appear
in IET Computer Vision.
Matlab code: download
here
The proposed SOAMST algorithm
is compared with the Mean Shift Tracking with a fixed scale [1], the Adaptive
Scale algorithm [1], and the EM-shift algorithm [2].
Experiment 1 ¨C The synthetic
ellipse sequence
|
Fixed Scale |
|
|
Adaptive scale |
|
|
EM-shift |
|
|
SOAMST |
Experiment 2 ¨C Palm
sequence
|
Fixed Scale |
|
|
Adaptive scale |
|
|
EM-shift |
|
|
SOAMST |
Experiment 3 ¨C Car
sequence
|
Fixed Scale |
|
|
Adaptive scale |
|
|
EM-shift |
|
|
SOAMST |
Experiment 4 ¨C Walking man
sequence
|
Fixed Scale |
|
|
Adaptive scale |
|
|
EM-shift |
|
|
SOAMST |
Reference
[1]
D. Comaniciu, V. Ramesh and P. Meer,
¡°Kernel-Based Object Tracking,¡± IEEE Trans. Pattern Anal. Machine Intell.,
vol.25, no. 5, pp. 564-577, May, 2003.
[2]
Z. Zivkovic and B. Kröse, ¡°An
EM-like Algorithm for Color-Histogram-Based Object Tracking,¡± In Proc. IEEE Conf. on Computer Vision and
Pattern Recognition, Washington, D.C., USA, vol. I, pp. 798-803, 2004.