A measurement fusion algorithm of active and passive sensors based on angle association for multi-target tracking
Yongquan Zhang,Aomen Shang,Wenbo Zhang,Zekun Liu,Zhibin Li,Hongbing Ji,Zhenzhen Su
DOI: https://doi.org/10.1016/j.inffus.2024.102267
IF: 18.6
2024-01-26
Information Fusion
Abstract:Multi-target tracking among different types of sensors is facing great challenge in fully utilizing various types of measurements. To this end, this paper presents a measurement fusion algorithm of single active and multi-passive sensors (SAMPS) based on angle association (AA), named SAMPS-AA algorithm, for multi-target tracking. Firstly, in order to narrow down the association range, the common angle measurements of two types of sensors are extracted by the proposed effective screening algorithm. Then, an exclusion strategy of wrong association groups is developed by building statistics, which is based on angle measurements. Subsequently, coordinates of fused measurements are obtained by angle association, based on least squares (LS). Finally, another exclusion strategy of wrong measurement points is proposed via measurement characteristics of active sensor. Experimental results indicate that the proposed SAMPS-AA algorithm can fully combine advantages of these two types of sensors, effectively exclude as many wrong association groups as possible, efficiently reduce the computational complexity, and obviously improve the tracking accuracy.
computer science, artificial intelligence, theory & methods