Mixed Traffic Flow Modeling Solution on Digital Map

This paper presents an advanced mixed traffic flow modeling solution utilizing deep learning and computer vision techniques integrated with digital mapping technologies. The proposed system employs YOLOv8 to detect and classify various traffic participants, including cars, trucks,… DeepSORT for multi-object tracking. OpenCV, along with our specialized algorithms, is utilized to estimate the speed of each detected object, recognize vehicle license plates, and determine the GPS location of each participant based on the camera's geolocation. These data points are then used to simulate real-time traffic flow on a digital map. The system's effectiveness is validated through a series of experiments, demonstrating its potential for digitizing traffic flow on a digital map, thereby enhancing traffic management and urban planning.