Vehicle tracking in multi- sensor networks by fusing data in particle filter framework

In this paper we propose a multi sensor tracking method. Tracking is done independently for each view. Fusing several cues including color, edge, texture and motion constrained by structure of environment is used in a novel way. Fusion of features in particle filter framework helps to achieve an accurate tracking algorithm in single view. The results of individual image planes are projected to the ground plane using homography relation. The similarity of the projected locations with the reference model and minimum variance estimate are two key points to evaluate the total location of the target. Experimental results show the robustness and accuracy of the proposed method.