Moving object tracking in support of unmanned vehicle opeartion

Multi-target tacking (MTT) has become an increasingly important research topic because of its various applications. One of the most recent and challenging implementations is to use MTT in road traffic so as to achieve autonomy for automotive applications from basic driver assistance level to full automation such as unmanned vehicles. This paper presents a scenario of tracking multiple moving objects in a traffic crossroad with an immobile unmanned vehicle. We introduce a Reid's based Multiple Hypothesis Tracking (MHT) data association filter to solve the problem of multi-target tracking in a cluttered environment. Murty's algorithm is used to find the M-best solutions in the assignment problem involved in MHT and N-scan pruning scheme is implemented to efficiently reduce the number of redundant hypotheses. Through simulation results it is verified that MHT based MTT technique can be effectively applied in unmanned vehicle operation.