Airborne multiple emitter tracking by fusing heterogeneous bearing data

Abstract In this work, we focus on the task of localizing and tracking multiple non-cooperative objects by a passive antenna array and an optical sensor. Both sensor systems are mounted on an unmanned aerial vehicle and collect bearing measurements from objects, whereby the number of the latter is unknown. Furthermore, imperfect detection of the optical sensor leads to false targets and clutter measurements. For localization and tracking, the imprecise but unique bearing data collected from the antenna array has to be correlated with the precise but ambiguous bearing data collected from the optical system. To avoid the computationally complex data association problem and dealing with a huge amount of data, an iterative approach is preferable, which only considers the sensor data from the current instant of time for updating each object state. This is realized by a Monte Carlo realization of an intensity filter. Firstly, the antenna array determines the direction of arrival of an emitter, and then the optical sensor detects possibly emitting targets in the corresponding area. Secondly, the bearing data of both sensors are fused in a single localization and tracking framework. In the proposed framework, the decision which object in the optical field of view is an emitter is made automatically. Experiments on simulated and real world data show that the proposed method is capable to localize and track multiple emitters using bearing data from heterogeneous sensors. The fusion of both sensor data types achieves a much higher accuracy in comparison to a system using only an antenna array.