Probabilistic Data Association with Amplitude Information versus the Strongest Neighbor Filter
Lisa M. Ehrman, William Dale Blair · 2007
Prior work on the probabilistic data association filter (PDAF) with amplitude information acknowledges that the strongest measurement receives preference regardless of its location in the track gate. On the other hand, if the measurements have relatively equal amplitudes, then assignment decisions are based on kinematics. Aside from those differences, prior work indicates that the PDAF with amplitude information operates the same as the standard PDAF. However, the literature understates the impact of the amplitude information on the PDAF. This paper analytically derives the decision regions (e.g., the regions for which the strongest measurement receives a larger association probability) for measurement-to-track association using a PDAF with amplitude information. The ratio of association probabilities corresponding to two measurements is also considered. The analysis shows that the PDAF with amplitude information often gives the strongest target an association probability approaching one. Thus, including an amplitude term frequently reduces the PDAF to a strongest neighbor filter.