Asynchronous Multi-Sensor Fusion Multi-Target Tracking Method
Song Liu, Han Shen-Tu, Huajie Chen, Dongliang Peng, Yifang Shi · 2018
The paper addresses the multi-target tracking problem for the asynchronous sensors system. As the performance of single sensor multi-target tracking method will degenerate in complicated environment, an asynchronous multi-sensor fusion algorithm based on Gaussian mixture probability hypothesis density (GM-PHD) filter is proposed. First, we construct a multi-sensor fusion framework for the GM-PHD filter. Then, to solve the data synchronization problem, we propose a time registration method based on state extrapolation. At last, we construct an improved covariance intersection method to fuse the posterior estimates. Simulation results show that, compared with the single-sensor GM-PHD algorithm, the proposed algorithm is more accurate and robust.