Novel Implementation of Track-Oriented Multiple Hypothesis Tracking Algorithm

It is widely accepted that modern computational capabilities have made the application of Multiple hypothesis tracking (MHT) feasible for a wide variety of applications. However, even in typical expected scenarios, periods of unusually high target or clutter density may occur that stress the ability of MHT to operate in real-time and under the constraints of limited computer memory. The most computing burden in MHT is the best global hypothesis formation. This paper establishes a solution tree and introduces the branch and bound strategy for the best global hypothesis formation. Then, a novel MHT algorithm which can be applied in practical radar implementations is proposed. The algorithm is illustrated with examples of simulated missile defense scenarios and a target tracking scenario with real radar data. The experiment results indicate that the algorithm is valid.