Improved MeMBer filter with modeling of spurious targets

The cardinality balanced multi-target multi-Bernoulli (CBMeMBer) filter removes the bias in the expected cardinality observed in the multi-target multi-Bernoulli (MeMBer) data update step. In this paper, a filter that offers a new statistical framework for the MeMBer data update step is proposed. Unlike the CBMeMBer filter, the proposed filter removes the positive bias by distinguishing spurious targets from actual targets in the MeMBer filter. To do this, the multi-target distribution of the multi-Bernoulli RFS is extended to model spurious targets arising from legacy tracks with high probabilities of existence. Simulation results are presented to demonstrate the effectiveness of the proposed algorithm.