Sensor fusion for target track maintenance with multiple UAVs based on Bayesian filtering method and Hospitability Map

In this paper, the track maintenance problem for tracking ground targets with Uninhabited Air Vehicles (UAVs) is studied. Different from traditional target tracking problems, each UAV here can only provide the information inside its own footprint, which is relatively small compared with the whole area of interest. Therefore, when the resources of UAVs are limited, the target information from the UAVs is both spatially and temporally incomplete. To deal with this problem, this paper proposes a generic multi-sensor fusion algorithm for target track maintenance with multiple UAVs. Track maintenance is achieved based on the Bayesian filtering method incorporated with the Hospitability Map, which is used to represent the terrain-dependent maneuvering behavior of ground targets. Simulation results are presented to demonstrate the capability of our approach in maintaining the tracks of targets even with intermittent and regional detection reports from UAVs.