Advanced Technique based on Nearest Neighbor for Tracking Closed Spaced Targets in Clutter

The In this paper, a new technique named optimum nearest neighbor data
association (ONNDA) is proposed to overcome the tracking issue of closed spaced moving
targets in dense clutter environment. The proposed algorithm detects the measurements that
represent the valid targets from all measurements in the cluttered gate. A new virtual gate is
assigned to the detected valid measurements. The center of this gate is represented by the
last point of the tracked target position. In this new gate the nearest neighbor data
association algorithm is used to select the true measurement that represent the moving
target. The ONNDA detects the candidate measurement with the lowest probability of
error, increases the data association performance compared to nearest neighbor (NN) filter,
and detects the closed moving targets in more background clutter. Simulation results show
the effectiveness and better performance when compared to conventional algorithm as
NNKF.