Video Tracking Based on Template Matching and Particle Filter
Shinfeng D. Lin, Ting-Yi Chen · 2018
In recent years, object tracking is still a challenging problem although many approaches have been successfully proposed. We propose a video tracking based on template matching and particle filter, for solving some issues in object tracking. The proposed method includes template matching and particles weighting. The object can be successfully tracked by template matching, except for some challenging sequences. To compensate for template matching, we exploit particle filter with Speeded Up Robust Features (SURF) to repair the failed tracking. Experimental results of the effectiveness and robustness are demonstrated, and the comparative performance is also shown.