Multitarget tracking in clutter using two-scan data association algorithm and fixed-lag smoothing

Abstract

The performance of multitarget tracking in clutter can be improved by using higher accuracy data association algorithm and more efficient state estimation method. This paper presents a novel fixed-lag TSDA-AI smoothing algorithm, which enhances the accuracy of state estimation by using fixed-lag smoothing, and improves the probability of real “measurement-target” combination by using two-scan measurements and the associated amplitude feature in the data association method. Its efficiency has been confirmed by computer simulations.