Road-Map Aided Gaussian Mixture Labeled Multi-Bernoulli Filter for Ground Multi- Target Tracking

Abstract

Ground multi-target tracking (MTT) is one of the core tasks of airborne ground moving target indicator (GMTI) radar and automotive radar. However, ground MTT still remains a challenging issue especially in complex traffic scenarios, since it often suffers from high clutter, dense targets, low visibility, etc. To enhance the tracking performance for ground multiple targets, in this paper, we present a comprehensive solution named road-map aided Gaussian mixture labeled multi-Bernoulli filter (RA-GMLMB) filter, which incorporates road-map information into the labeled multi-Bernoulli (LMB) filter. Specifically, we first propose a hybrid circular arc and line segments approximation approach to extract road-map information, which alleviates approximation errors in the procedure of road-map approximation. Then, we deduce the RA-GMLMB filter by integrating the extracted road-map information into the LMB filter with Gaussian mixture implementation. Simulation experiments are conducted and experimental results show that the proposed RA-GMLMB filter outperforms the state-of-the-art methods.