Robust joint association and registration under large sensor bias
Na Ni, Qi Jiang, Huafeng Mao, Jichuan Zhang, Rui Wang, Cheng Hu · IET conference proceedings. · 2024
Radar observation of group targets has recently received much attention in the flight mechanism research and many other applications. Group targets are usually closely spaced with lots of missed detections and false alarms, making it difficult for spatial reconstruction. Multisensor systems make use of data from multiple radars to provide more accurate measurements and robust tracks than a single radar, playing an important role in the group targets observation. In the data fusion processes, multisensor measurements are associated after transformed into a global coordinate system. However, large sensor bias makes it difficult to associate measurements of closely spaced targets given outliers, and nonideal association further increase the sensor bias estimation error. In this paper, we proposed a robust joint association and registration method to simultaneously acquire sensor bias estimates and association results. A nonlinear least median of squares estimator is used to achieve better accuracy and robustness of sensor bias estimation. Simulation results show the better bias estimation and association performance of the proposed method compared to other algorithms.