论文部分内容阅读
A motion filtering method is proposed in this paper,which takes into consideration the exist-ence of certain system constraints with respect to the amount of the corrective rotational and transla-tional motions that can be applied on each video frame for stabilization.The interdependence be-tween rotational and translational constraints is considered,and a constraint unscented Kalman filter (CUKF)algorithm is utilized to achieve a smoothly stabilized motion.The method is applied to sta-bilize each video frame,and the results reveal that the proposed approach improves the stabiliza-tion performance in comparison with the trivial approach.