Tutor: Tianran Wang
The Kalman Filter is widely used in localization, navigation, etc., where the measurements contain noises and other inaccuracies. The Extended Kalman Filter is applied when the mathematical model is nonlinear.
The algorithm has two main processes. In the prediction process, the Kalman filter makes an estimation of the state along with the uncertainties through the mathematical model. In the update process, the Kalman filter updates the state using the measurements from the sensors, whereby the certainty is increased.