The Kalman filter provides a general solution to the recursive, minimum mean-square state estimation problem within the class of linear estimators, assuming that the dynamics of the object of interest and measurement noise are accurately modeled. As applied to scenarios where a radar or another active sensor is used to track objects, this filter estimates the object’s state (e.g., position and velocity) at some time, usually the predicted time of the next observation, and then updates that estimate using noisy measurements. It also offers an estimate of the object’s tracking error statistics through the state error-covariance matrix.
|