Higher-Order Moment Based Target Tracking
Liping Song, Hongbing Ji · 2006
Being different from Kalman filter, higher-order moment based target tracking provides a new solution to a target tracking. It is unnecessary to compute the Kalman gain matrix and the filtered corresponding covariance matrix, so as to avoid the determination of the initial filtered corresponding covariance matrix which is difficult sometimes. The new method presented in this paper predicts and estimates the state of the target only through the third-order moment of the measurement vector. It needs not any knowledge about the measurement noise. The result deduced is very simple and is convenient for engineering.