A high estimate accuracy fusion method for multisensor tracking association

A high estimate accuracy fusion method for multisensor tracking association is presented. Based on the analysis of the state fusion and measurement fusion methods, the estimating effects of the two methods are compared. Considering the measurement covariance's influence on the fusion result, we assign the elements of the measurement covariance matrix to the measurements from different sensors as their weights, and combine the weighted measurements to form the new measurement which is used in the Kalman filter. The simulation shows the new method's estimate accuracy is much higher than the state and measurement fusion methods.