Asynchronous track fusion in a multi-scale sensor environment
Yue Shi, Yuhan Dong, Xiuming Shan · 2004
Track fusion is an important and attractive part in multisensor multitarget tracking for its robustness and flexibility compared with measurement fusion. Much research on track fusion has been done under the common assumption that the sensors in system are synchronous (with no communication delays) or have similar sensor accuracies. Unfortunately, in many multisensor systems, situations often fall short of the above assumption due to the mobility of targets and/or sensors, the communication restrictions, and the type and resource limitations of the individual sensors. The aim of this paper is to propose an asynchronous approach for track fusion in a multi-scale sensor environment, which we refer to as asynchronous hierarchical track fusion (AHTF). It is shown that the proposed AHTF approach provides a solution to combining tracks reported by multi-scale sensors, which are dissimilar in their data rates and resolving capacities. Simulations in a typical maritime surveillance scenario are presented to validate the performance of the proposed approach.