Estimation with Applications to Tracking and Navigation
Estimation with Applications to Tracking and Navigation
An Introduction to the Kalman Filter
An Introduction to the Kalman Filter
| Hi-index | 0.00 |
The goal of a tracking system is to follow the target trajectory using useful information about the target's state from sensor observations. In this kind of applications, where we do not know the target motion a priori, is very important to have a good model of the target kinematic and an estimation filter for control of the sensor position. It is also very useful to have a robust filter working successfully in real dynamic environment. To determine this kind of robustness the goal of this paper is to compare the performance of four estimation filters for tracking objects on digital images: Kalman filter (KF), Extended Kalman filter (EKF), Unscented Kalman filter (UKF) and Interacting Multiple Model (IMM) estimator using different target kinematic models and different image sequences. These sequences include six real and seven simulated scenes. We compared the performance of the filters using the root mean square (RMS) position error, for both filtering and prediction filter modes.