From target tracking to targeting track: A data-driven yet analytical approach to joint target detection and tracking

This paper addresses the problem of real-time detection and tracking a non-cooperative target in challenging scenario with almost no a-priori information about birth, death, dynamics probability. Furthermore, there are false missing data at an unknown yet low rate measurements. The only given advance is target-measurement model constraint that more than one scenario. To solve these challenges, we movement by using polynomial trajectory function time (T-FoT), which aims to estimate continuous-time rather series discrete-time point estimates as done most existing filters/trackers. Data-driven T-FoT initiation termination strategies proposed for identifying (re-)appearance disappearance target. During existence target, real measurements distinguished from clutter if indeed exists detected, order update each scan design least-squares estimator. Overall, our approach Markov-free, data-driven analytical. Simulations either linear or nonlinear systems conducted demonstrate effectiveness comparison Bayes optimal Bernoulli filters. results show comparable perfectly-modeled filters, even outperforms them some cases while requiring much less computing faster.