CBMeMBer Filter for Extended Target Tracking Based on Binomial Measurement Number Model

Targets that give rise to multiple measurements for each scan in high-resolution sensors are defined as extended targets. In general, existing algorithms based on the random finite set (RFS) theory assume that the number of measurements generated by an extended target follows a Poisson distribution; however, this assumption has been found to be inaccurate and inconsistent with actual situations. To address this problem, an extended target cardinality balanced multi-target multi-Bernoulli (ET-CBMeMBer) filter based on a binomial measurement model is proposed. Firstly, it is assumed that each extended target's measurement number is binomial distributed. Then, its updated equations are analytically derived and relevant proofs are provided. Finally, simulated results illustrate the proposed filter's effectiveness and superior tracking performance compared to the Poisson ET-CBMeMBer filter.