Track initiation with ant colony optimization

Since the issue of track initiation belongs to the NP-hard problem in the bearings-only multi-sensor-multi-target tracking system, a novel proposed track initiation technique is proposed in this paper. The proposed track initiation technique is based upon an ant colony optimization (ACO) algorithm, a kind of heuristic optimization method. Observing that each target is of the characteristic of uniform rectilinear motion, we develop a new cost function derived from the thought of Hough transform. Numerical simulation results show that the proposed ACO-based track initiation method not only meets the requirement of real time, but also performs better than other traditional techniques, especially in the scenario that all targets move in parallel.