Development of AI-Based Vehicle Detection and Tracking System for C-ITS Application

There are various means of monitoring traffic situations on roads. Due to the rise artificial intelligence (AI) based image processing technology, there is a growing interest in developing systems using camera vision data. This study provides method for deriving information installed at an intersection improve system The uses deep-learning-based approach (YOLOv4) vehicle detection and type classification. Lane-by-lane trajectories estimated by matching detected locations with high-definition map (HD map). Based trajectories, volumes each lane-by-lane traveling direction queue lengths lane estimated. performance proposed was tested thousands samples according five different evaluation criteria: rate, classification, trajectory prediction, volume estimation, length estimation. results show 99% less than 20% errors classifying types estimating travel volume, which reasonable. Hence, this shows feasibility collecting detailed intersection. combining AI HD techniques main contribution study, high chance improving current systems.