Track Split Smoothing for Target Tracking in Clutter

The proposed integrated track split (ITS) smoothing utilize smoothing data association algorithm for tracking a target. We implemented two independent ITS filters; forward running ITS (fITS) filter and backward running ITS (bITS) filter. The novelty of the proposed algorithm is that the backward multi-track component predictions are applied to each forward track component prediction to produce multiple information fusion component predictions based on the data association technique. The information fusion state predictions are applied to all available measurements received in current scan to smooth track state estimations and target existence probabilities. A new technique is developed which applies smoothing estimates to compute the fITS estimate in the current scan. This efficiently leads forward path track to track a target in heavy clutter. The algorithm is known as fixed-interval smoothing ITS (FIsITS). The numerical simulation verifies the false track discrimination (FTD) capability of the FIsITS.