Tracking using Converted Measurement Kalman filter through improved algorithm in the missed detections scenario

Estimating the position of a moving target in the polar frame of reference has been a major problem in the conventional tracking systems. The commonly used sensor equipment provides the position of target in polar coordinates i.e. in range and azimuth (or bearing) angle with respect to the sensor location. The use of simple Kalman filter increases the error in this case. For more accurate tracking, the Converted Measurement Kalman filter (CMKF) is used which can account for the inaccuracies in tracking using polar coordinates. Simulations were performed tracking the target with CMKF. It does not proffer well in missed detection and false alarm scenarios. So the tracking was improved by associating Global nearest Neighbour algorithm (GNN) algorithm with the CMKF. Later the simulations depict the inconsistency of GNN based CMKF in a dense missed detection and false alarm scenarios. So an improved algorithm in track estimation is used to solve the dense missed detections or continuous false alarm problem. Thus, through simulation, it was realized that GNN based improved CMKF is able to give a better tracking. Finally Wiener filtering is used to smoothen the tracking.