A Dirac Delta mixture-based Random Finite Set filter

A Random Finite Set (RFS) based multi-target filter using a mixture of multi-object Dirac Delta and Poisson RFSs is proposed. The resulting distribution is closed under the Chapman-Kolmogorov equation while also being a conjugate prior to the “natural” multi-target likelihood function. A filtering algorithm is presented which efficiently extracts the highest weight components of the complete mixture distribution. Results show that the proposed method outperforms the Probability Hypothesis Density filter and the Cardinality Balanced multi- Bernoulli filter RFS-based methods in simulated environments.