Track-to-track association and ambiguity management in the presence of sensor bias

摘要: The track-to-track association problem is to determine the pairing of sensor-level tracks that correspond same true target from which sensorlevel originated. This crucial for multisensor data fusion and complicated by presence individual sensor biases, random errors, false tracks, missed tracks. A popular approach performing between two systems jointly optimize a posteriori relative bias estimate sensors likelihood association. Algorithms solve this typically generate K best bias-association hypotheses corresponding likelihoods. In paper, we extend above in ways. First, derive closed-form expression computing “pure” likelihoods, as opposed likelihoods are weighted unique estimate. Second, present an alternative formulation solely with respect These results facilitate what commonly known system-level track ambiguity management.