Based-APF human motion tracking from monocular videos

In this paper, a novel method based on adaptive particle filter (APF) for tracking human motion in monocular videos is proposed. With an initial human skeleton joint point template, we use the probability density propagation of the particle filers through the model. This algorithm can automatically deal with tracking issues such as occlusion and auto-occlusion. Experimental results from 20 classes monocular videos show that the proposed method is robust and that the tracking results are good.