Pedestrian tracking algorithm based on Kalman filter and partial mean-shift tracking

In recent years, the object of study in the science has gradually transmitted from vehicle to pedestrian and has an increasing requirement of the accuracy. In aspect of pedestrian tracking, this thesis introduces a new tracking unit based on the fact that in most cases, the occlusions of the pedestrians are partial and using it in multiple pedestrians tracking. Results shows that the tracking unit has a better performance in case that pedestrians has frequent occlusions and reduce the possibility of miss tracking or error tracking.