Histogram PMHT with particles
Samuel J. Davey · Adelaide Research & Scholarship (AR&S) (University of Adelaide) · 2011
The Histogram Probabilistic Multi-Hypothesis Tracker (H-PMHT) is a parametric mixture-fitting approach to track-before-detect. Recent comparisons have shown that it can give performance close to numerical approximations to the optimal Bayesian filter at a fraction of the computation cost. The derivation of H-PMHT makes no explicit assumption about the target process model or the sensor point spread function: these details are dictated by the application. However, only linear Gaussian implementations have been used in the literature and there is a growing misconception that H-PMHT requires linear Gaussian models. This paper considers the implementation of H-PMHT for non-linear non-Gaussian problems.