Computational performance improvements of multiple hypothesis tracking algorithm
Kwiatkowski Marek, Mioslaw Sankowski, Dawid Lukwinski · 2014
Multiple hypothesis tracking (MHT) is considered to be the preferred solution for track formation and target tracking in complex environment. Main difficulty with MHT is it's high computing power requirements. This paper deals with several aspects of computational performance of this algorithm. Improvements like space division, multithread processing and fast hypotheses generation techniques are considered.