3D Motion Estimation of Hand Based on Feature Points

A new method for estimating 3D motion and structure of finger from 2D image sequence is proposed in this paper. This problem is challenging not only due to the correspondence problem but also the lack of depth cues in 2D image sequence. In this paper, affine motion is chosen as a suitable model for finger's motion. However, the motion model can also be extended and not necessarily be affine. Analysis of intensity images is first used to find the correspondence for each feature point. Then affine motion model is utilized to find the structure and motion parameters. Extensive analysis has been done on how defining appropriate constraints which are necessary for achieving convergence. Finally, experimental results are presented. The results are very encouraging and have many potential applications especially in such fields as gesture recognition, virtual reality, animation and motion tracking