Introducing Contextual Information in Multisensor Tracking Algorithms

Perception systems used in fields like robotics and aircraft tracking have recently been developing toward the use of combined sensors of different kinds. Nevertheless, the measurement management of such systems has to be adapted to the sensors used, and particularly to all modifications in the utilization conditions. For this task we propose here to use a set of heuristics to modify the operation of a Kalman filter dedicated to air vehicle tracking.