Multi-Target Multi-Sensor Tracking Problems: A General Bayesian Approach

Based upon a general probabilistic target/sensor model and the Bayes´ rule, a general solution to the multi-target tracking problem is derived. A less general but more implementable class of multi-target tracking algorithms is shown with an additional set of independence assumptions. Existing algorithms are then related to the general formulation. Hypothesis management (reduction) and other implementation issues are briefly discussed.