Non parametric target tracking in non uniform clutter

Target tracking algorithms have to operate in an environment of uncertain measurement origin, in the presence of randomly detected target measurements as well as clutter measurements from unwanted random scatterers. Most data association target tracking algorithms incorporate a measure of clutter density. Quite often the clutter density is non-uniform, especially in ground target tracking, or when the radar beam "touches" the ground. Non-parametric target tracking assume a priori unknown clutter density, however the assumption is that it is uniform within the selection gate. Complex selection gates, which are created for a number of multi-scan and/or multi target tracking algorithms, can encompass a large volume where the clutter density will vary widely. We propose an alternative clutter density estimation scheme for non-parametric target tracking, which does not assume uniform clutter within the selection gate. Instead, clutter density is estimated on a measurement by measurement basis, prior to a posteriori target track update. The scheme is adaptive and has no bias in estimated inverse of clutter density at the measurements coordinates. A simulation study shows the benefits of adaptive clutter estimation based on single target tracking algorithms in an environment of heavy and non-uniform clutter.