Generalized Labeled Multi-Bernoulli Track Initiation Algorithm Based on Multi-Hypothesis Tracking

The generalized labeled multi-bernoulli filtering (GLMB) algorithm requires prior information on birth probabilities and initial positions, but these are unknown in practical engineering. In this paper, a track initiation algorithm based on multiple hypothesis tracking (MHT) is proposed. The algorithm uses multi-hypothesis tracking filtering and real-time measurement information to form an experimental track as a possible birth target initial point. The filtering results are smoothed by Rauch-Tung-Striebel (RTS), and the processed track is taken as the birth information of the generalized labeled multi-bernoulli filtering algorithm. The effectiveness of the algorithm in this paper was verified through simulation comparisons.