ABSTRACT
Accumulated state density (ASD) provides a unified treatment of filtering and retrodiction insofar as by marginalizing them appropriately, the standard filtering and retrodiction densities are obtained. ASDs are useful in tracking applications, where out-of-sequence (OoS) measurements are to be processed, that is, when sensor data do not arrive in temporal order, in which they have been produced. This chapter discusses why ASDs provide an exact solution to the track-to-track fusion (T2TF) problem. It summarizes basic facts on the Bayesian tracking paradigm. The chapter presents the notion of an ASD along with a discussion of closed-formulae for the parameters of the ASD in the case, where Kalman filtering can be applied to tracking. It also discusses the role of ADS within the probabilistic multiple hypothesis tracking (PMHT) framework. A Bayesian tracking algorithm is an iterative updating scheme for calculating conditional probability density functions p(xl|Zk) that represent all available knowledge on the object states xl at discrete instants of time tl.
