Multiple target tracking with quantized measurements: A standard Bayesian approach

This paper shows how the use of standard Bayesian likelihood functions and particle filters provides a simple and general approach to tracking with quantized measurements. We demonstrate this result with two examples. The first involves a maneuvering target and the second multiple targets. In both cases the motion model for the target(s) is non-Gaussian and the quantized measurements correspond to a set of possible locations composed of disjoint regions in 2-space. For the multiple target example, we apply a nonlinear particle filter version of JPDA to perform the data association and tracking.