Research On Targent Tracking Based On Metabolic Model And LK Tracking

The problem of target tracking drift is usually caused by the occlusion and the perspective change in complex environment. To solve the problem, this paper proposes a new method based on the combination of the metabolic model and the Lucas-Kanade (LK) registration algorithm. The proposed method realizes the effective target tracking in the state of complex motion through a two-stage progressive tracking mode. The two-stage progressive tracking mode utilizes the dynamic template and the quasi-static template of the metabolic model for target tracking, respectively. The dynamic template of the metabolic model is updated in real time, while the quasi-static template of the metabolic model is updated periodically. The metabolism from old characteristics to new characteristics of target is realized by the utilization of the relationship between the target motion information and the template change information. Experiment results show that the proposed method has good qualities in terms of robustness and real-time property.