Publisher's Synopsis
In recent years, the Multiple Hypothesis Tracker has gained acceptance as means of handling targets in ameasurement-clutter environment. MHT algorithms rely on Gaussian mixture representations of a target's current stateestimate, and the number of components within these mixtures grows exponentially with each successive sensor scan.Previous research into techniques that limit the growth of Gaussian mixture components proved that the Integral SquareError cost-function-based algorithm performs well in this role. Also, multiple-model adaptive algorithms have been shownto handle poorly-known target dynamics or targets that exhibit a large range of maneuverability over time with excellentresults.
