Group target data association with partially overlapping observation areas

Group targets data association is a key part of group targets analysis through multisensor information fusion. Traditional data association methods using group centroids are no longer applicable when the observation areas of multiple sensors partially overlap. Aiming at the problem a new group targets data association algorithm is proposed in this paper. Firstly, the local structures centered on every point within the point pattern which represents the group target data are constructed based on its Delaunay graph. A pair of reference subsets is found from two point patterns to be associated via local structure matching. Then the two point patterns are aligned by the transformation parameter estimated using the reference subsets. Finally the various group targets data are associated once the aligned data have been matched. Simulation experiments show that the proposed algorithm is an effective solution for the group targets data association problem of multiple sensors with their observation areas partially overlapped.