Distributed multi-camera object tracking with Bayesian Inference

A novel algorithm is proposed to perform object tracking with multiple cameras in the Bayesian Inference framework. The key contribution is the exploitation of Bayesian network to fuse spatial-temporal position and object template feature in multiple cameras. Firstly, Bayesian network is used to model the multiple static cameras´ tracking system. Then, the high-dimensional joint posterior is propagated spatiotemporally. Finally, the estimation of the target location in each camera view is achieved by using sequential Monte Carlo Approximation. The robust tracking algorithm efficiently fuses information from different views and is capable of dealing with partial and full occlusion. Besides, the distributed tracking algorithm is implemented on OMAP3530 platform. Both qualitative and quantitative experiments have demonstrated the effectiveness and robustness of the proposed algorithm.