The Multi-target Tracking Algorithm Based on the MM-GLMB Filter with Doppler Information

Abstract

The traditional multi-target tracking (MTT) algorithm only uses the position information of the target for target tracking, which is prone to more false alarms when dealing with the multi-target tracking problem in the dense clutter environment, and it will increase the complexity and computation of data processing, and even generate the false tracks or lose tracks. In addition, the threshold of tracking gate of the traditional MTT algorithm is fixed, and the probability of target miss detection or false alarm is greatly increased in dense clutter environment. Aiming at the above problems, a tracker is built in order to improve the tracking performance in this paper. Firstly, Doppler information (or radial velocity) is introduced based on the multiple models generalized labeled multi-Bernoulli (MM-GLMB) filter under the consideration of the correlation between range and radial velocity. Secondly, the multi-dimension tracking gate is constructed with the information of target position and velocity in polar coordinate system, and the threshold of the multi-dimension tracking gate is adjusted through a maximum likelihood adaptive method. Finally, the state of target is updated using position measurements followed by sequential updates using Doppler measurements for make full use of the Doppler information. The results of the experimental simulation show that introducing the Doppler information of the target based on MM-GLMB in the dense clutter environment and the processing of the measurements using the adaptive tracking gate can effectively suppress the clutter and significantly improve the tracking performance.