Consensus protocols for distributed tracking in wireless camera networks

Consensus-based target tracking in camera networks faces three major problems: non-linearity in the measurement model, temporary lack of measurements (naivety) due to the limited field of view (FOV) and redundancy in the iterative exchange of information. In this paper we propose two consensus-based distributed algorithms for non-linear systems using the Extended Information Filter as underlying filter to handle the non-linearity in the camera measurement model. The first algorithm is an Extended Information Consensus Filter (EICF) that overcomes the effect of naivety and non-linearity without requiring knowledge of other nodes in the network. The second algorithm is an Extended Information Weighted Consensus Filter (EIWCF) that overcomes all the three major problems (naivety, redundancy and non-linearity) but requires knowledge of the number of cameras (Nc) in the network. The basic principle of these algorithms is weighting node estimates based on their covariance information. When Nc is not available, EICF can be used at the cost of not handling the redundancy problem. Simulations with highly maneuvering targets show that the two proposed distributed non-linear consensus filters outperform the related state of the art by achieving higher accuracy and faster convergence to the centralised estimates computed by simultaneously considering the information from all the nodes.