Multiple target tracking with possibly merged measurements modeled by point processes

Generally, sensor resolutions are finite, and hence any measurement may originate from two or more objects. Recent advances in sensor signal processing technology have made it possible to estimate the number of objects that may originate from a given detection for a certain type of sensor. With these new methods, it may be possible to associate each, possibly merged, measurement with state components of an unknown number of objects from which the measurement originates. One suitable mathematical model for an unknown number of object states is a finite point process that can be defined by a family of permutable (symmetric) probability distributions together with the probability of the number of the objects. In this paper we will show how general multiple hypothesis tracking systems, usually designed with the assumption that measurements are all resolved, may be extended to process possibly merged measurements modeled by point processes.