A window cumulative normalized distance based phase optimization track association model

Abstract

A novel approach is proposed, which is able to overcome several shortcomings existing in the typical distributed multi-sensor track association (MSTA) model-the sequential minimum normalized distance nearest neighbor (SMNDNN) correlation and the sequential minimum mean square error (SMMSE) fusion. By considering the mutual dependency in the track estimation errors between different sensors and using the window cumulative normalized distance (WCND) technique, this phase optimization algorithm can guarantee stability in dealing with track association, especially in a dense track environment. The experiments demonstrate that our model can efficiently resolve the MSTA, the ad hoc dense track association problem.