Multiple signal tracking processes

To discern an object of interest in a complex environment, the situation assessment has to get the most out of the available discriminatory features that can be extracted from the signals generated by the various sensors for each resolution cell. To meet that need, a larger tracking method is proposed here that directly associates and filters identity likelihoods built on such features, in the framework of theory of evidence. This method offers a number of advantages deriving from the fact that it avoids local detection decisions; makes synergetic use of the ''classification'' and ''tracking'' functions; provides integrated, adapted processing of ambiguities when matching multiple observations; manages target signature uncertainties appropriately; and provides a suitable track validation procedure.