An Efficient Data Association Algorithm for Single Target Tracking with Multiple Centralized Heterogeneous Sensors

An efficient data association algorithm is presented to track a single target with multiple centralized heterogeneous sensors. The sharing parts of measurements from different sensors are used to accomplish the data association, which utilizes the 2 test match combined with Nearest Neighbor association strategy to depress the dummy measurements. Targetoriginated measurement matched pair will be retained and fused to update the target track. Unscented kalman filter is deployed to target state estimation for both the dynamic and measurement model are nonlinear. Simulations validate the algorithm and conclusions.