A modified probabilistic data association filter in a real clutter environment

A new multiple target tracking algorithm for tracking in clutter is developed. The new algorithm uses a modified Hough transform (HT) for track initiation and a directional probability data association (DPDA) filter for data association (DA). Two basic criteria are used to distinguish the tracks from clutter during the track initiation phase. Returns from only five radar scans are required. The DPDA incorporates the directional information and the association probabilities are estimated using both Mahalanobis distance and the track direction. This algorithm is designed for target tracking where the effect of clutter is considered to be severe. Simulated data as well as real radar data are used to evaluate the performance of this new tracking system.