3D motion estimation of head and shoulders in videophone sequences

In this paper, an approach for 3D motion estimation of head and shoulders of persons in videophone sequences is presented. Since head and shoulders are linked together by the neck, constraints for the motion of head and shoulders exist which are exploited to improve the motion estimation. In this paper, the human neck joint is modeled by a spherical joint between head and shoulders and its 3D position is calculated. Then, rotation and translation of the shoulders are estimated and propagated to the head. Finally, a rotation of the head around the neck joint is estimated. The presented approach is applied to the videophone sequences Akiyo and Claire. Compared with an approach without using a neck joint, anatomically correct positions of head and shoulders are estimated.