Data Fusion of Roadside Camera, LiDAR, and Millimeter-Wave Radar
Shijie Liu, Jianqing Wu, Bin Lv, Xinhao Pan, Xiaorun Wang
Abstract
Roadside sensor data fusion is an essential component of the vehicle-road cooperation system, thus effectively enhancing the interactive perception level among road targets. However, due to the complex road environment, occlusion, and other problems, the single sensor has low accuracy in the process of target tracking. How to realize the fusion of multisensor trajectory tracking data is the main problem to be solved at present. Therefore, a new multisensor data fusion method for roadside camera, LiDAR, and millimeter wave (mm-Wave) radar is proposed in this study. According to the change in reflection intensity caused by the shift of the LiDAR point cloud with the change in distance and the detection accuracy of mm-Wave radar used in this article, the weight parameters of LiDAR and mm-Wave radar in the fusion process are determined. Finally, the target missed detection rate and trajectory disconnected repair rate were customized, and experimental tests were conducted in five natural environments to verify the robustness of the proposed method.