Modified Mean-Shift for head tracking
Daeha Lee, Jaehong Kim, Joo-Chan Sohn · 2011
Mean-Shift algorithm has been applied mode seeking, image segmentation and object tracking. It is famous for fast convergence within limited boundary. But once exit that boundary, it fails to find correct region and position. We applied mean-shift algorithm in object tracking, and we suggest modified mean-shift tracking method. Using this modified version, we can track object without failing.