Track-to-track association using fuzzy membership function and clustering for distributed information fusion

In distributed information fusion application, developing an efficient track-to-track association approach becomes crucially important which may significantly benefit the sequent track-to-track fusion procedure. This paper proposes a novel track-to-track association method specialized for distributed multitarget tracking using more than two sensors. In order to mathematically interpret how probable that the two tracks from two different sensors are tracking the same targets, the fuzzy membership of the two tracks are calculated, whose value is between 0 and 1, with bigger values indicating higher probability that the two tracks originate from the same target. Based on the calculated fuzzy membership matrix, the clustering methodology is then utilized to pick out the group of tracks tracking the same target. The simulation results validate the efficiency and superiority over the existing approach.