A superimposed intensity multi-sensor GM-PHD filter for passive multi-target tracking

In order to address the problem of tracking multiple targets with multiple passive sensors in the presence of clutter and miss-detection, a new multi-sensor Gaussian mixture probability hypothesis density filter, together with an intensity superimposing technique, is proposed in this paper. First, a gate technique is used to choose valid sensors for each predicted Gaussian component. Second, each Gaussian component is updated with measurements from valid sensors, and the fusion intensity is obtained by adding all the local posterior intensities and the global miss-detection intensity. Last, a two-step extraction method is proposed to estimate the cardinality and states of the targets. Simulation results verify the effectiveness of the proposed method.