Bayesian target selection after group pattern distortion

The problem of uncertainty in measurement/target association is addressed. A group of N targets is observed at time t, and one of them is selected. At some later time t, the target group is again observed. The targets are assumed to move as a group, as well as individually, between observation time steps and so have a dependent motion model. The detection probability at t, is less than one and so not all targets may have been detected. There is also the possibility that not all measurements received at t, originate from targets. The problem is to estimate the state of the target originally selected, given the two sets of measurements. The novel feature of this analysis is the use of the dependent motion model