Moving target dynamic parameter detection method suitable for complex traffic scene

The invention relates to a moving target dynamic parameter detection method suitable for a complex traffic scene, and belongs to the technical field of intelligent traffic. The method comprises the following steps: S1, acquiring point cloud data of a traffic scene based on a millimeter wave radar, and acquiring image data of the traffic scene based on a visual camera; s2, unifying coordinate system planes: projecting point cloud data of a traffic scene to a plane where image data is located by adopting a space-time fusion model of a millimeter-wave radar and a visual camera; s3, determining a millimeter wave radar region of interest, namely screening radar target data from the point cloud data; and S4, carrying out target detection on the image data by adopting an improved YOLOv4 multi-layer parameter model compression network, and outputting a target category. According to the invention, the environment sensing capability of the vehicle in a complex traffic scene can be improved, the stability and reliabilit