A novel method for multi-target tracking
Songhao Zhu, Xinshuai Zhu, Zhuofan Li, Juanjuan Hu · 2014
This paper proposes a method to deal with the issue of multi-target tracking by considering continuous energy minimization. Specifically, observation model, appearance model, dynamic model, mutual exclusion model, trajectory persistence model and trajectory regulation model are integrated into an objective function; then, the gradient descent method is utilized to solve the constructed objective function to obtain the approximate minimum energy at every moment, and obtain the number and status of multi-target; finally, the intelligent extrapolation method based on the continuous energy minimization is utilized to achieve the final smoother and more continuous tracking trajectories. Experimental results on PETS 2009/2010 and TUD-Stadtmitte video database demonstrate that effectiveness and efficiency of the proposed MAP-HOR scheme.