Vision based aircraft tracking under LPC coordinate system

Implementations of the Bayes filtering for vision based aircraft tracking suffer fundamental difficulties in the Cartesian coordinate. This is because of the uncertainty of the noise model in the flight dynamic equation when the aircraft state is described in these coordinates. Recently, the Logarithmic Polar Coordinate (LPC) framework has been successfully applied in the bearings-only tracking context, with the ability of estimating the variance to range ratio (Brehard and Le Cadre [2004]). In this paper, we design an aircraft tracking method using LPC. This method can adapt with changes of the noise model in the flight dynamic equation. Hence, it is more robust than methods for vision based aircraft tracking using the Cartesian Coordinates.