Kalman Filter-based Multisensor Data Fusion

Kalman Filter-based Multisensor Data Fusion

In this paper, we evaluate the performance of two multisensor data fusion techniques for state estimation by applying Kalman filter. These methods are state vector fusion and measurement fusion. The comparisons of these methods are demonstrated for a target tracking problem and analysis is performed by means of the components of the error covariance matrix. According to environmental conditions, we should select one of the fusion architectures in order to fuse data obtained from respective sensors. The simulation results show that the measurement fusion methods generally have better state estimation performance over the state vector fusion methods.