H-PMHT with an unknown arbitrary target

Abstract

The Histogram Probabilistic Multi-Hypothesis Tracker (H-PMHT) is a parametric track-before-detect algorithm that has been shown to give good performance at a relatively low computation cost. The original algorithm assumes a known target signature and provides joint detection and tracking. A recent advance has allowed for the estimation of a time evolving Gaussian signature. This paper introduces a non-parametric method for learning an arbitrary target signature. The two methods are compared on Gaussian and non-Gaussian targets.