Optimal set integer programming algorithm for multiple maneuvering targets tracking in clutter

The aim of this paper is to solve the problems of multi- target tracking in clutter. Firstly, the data association of measurem ent-to-target is formulated as an integer programming pro. blern. Through using the linear programming (LP) based branch- and-bound method and adjusting the constraint conditions, an opt imal set integer programming (OSIP) algorithm is then proposed for tracking multiple non-maneuvering targets in clutter. For the case of maneuvering targets, this paper introduces the OSIP algor ithm into the filtering step of the interacting multiple model (1MM) algorithm resulting in the 1MM based on OSIP algorithm. Extensive Monte Carlo simulations show that the presented algorithms can obtain superior estimations even in the case of high density noises.