Effects of Measurement Uncertainties on Multitarget Tracking

Target tracking is the process of estimating the positional attributes of unknown targets using the measurement data from sensors like radar, LiDAR, camera of air, space, ground, maritime, underwater and autonomous systems. Target tracking involves finding the exact number of targets at any given time and identifying the state (e.g., location and velocity) of the targets over time in the platform or earth coordinates using the measurement data. An accurate measurement data will produce good track estimates. However, in a real-world scenario, none of the measurements are perfect due to several factors, like the resolution of the sensor, noise and inaccuracies originating from external influences. Uncertainties in the measurements will degrade the performance of any trackers. Ignoring some of the errors in the measurements for simplicity may lead to wrong spatial perception.