Enhancement of Data Separability in Multisensor-Multitarget Tracking Problems

In many current tracking problems, the tracking scenario may involve multiple targets whose states are to be estimated with observations from several sensors. The various aspects of the multitarget problems and the outlines of different solution techniques are given. In this paper improvements to a clustering analysis-based modular approach are made through the use of observation residuals. In such an approach, the multitarget problem is reduced to several single-target problems