Multisensor triplet Markov chains and theory of evidence (Q997027)

This paper presents Dempster-Shafer fusion methods in the context of different multisensor Markov models and applies these methods in some particular models such as Pairwise or Triple Markov Chains (PMC and TMC). A general presentation of stationary Hidden Markov chains and their generalizations PMC and TMC is provided. The main results concern the connection between PMC, TMC and the theory of evidence, through the DS fusion, introducing the Evidential Markov chains, for which a couple of results regarding the time required to compute the marginal distributions are found. Further results in this direction are developed in the particular case of PMC using evidential theory.