A multi-group target tracking algorithm based on probabilistic multi-hypothesis tracking

Aiming at the problem of tracking multi-group target with joint estimation of their motion state and shape, a multi-group target tracking algorithm based on probabilistic multi-hypothesis tracking (PMHT) and random hypersurface model (RHM) is proposed. The algorithm views the group target as a whole, models the extended shape of the group target with RHM, and then uses PMHT to estimate the state of the group target with the maximum a posteriori estimate. The algorithm addresses the challenge of accurately associating measurements with the group target and provides precise estimation of the complex shape of the group target. Finally, it verifies the effectiveness of the algorithm through simulation experiments.