Data Modeling and Architecture for a Machine Learning-Based Vehicle Counting and Classification System

Transportation difficulties have been the source of anxieties around the world which is caused by rising population and a rapid surge of automobiles at peak hours which results in traffic jams. In return, studies about control systems for intelligent traffic lights have been popular in recent years that they were able to show they are capable of governing traffic more efficiently compared to previous technologies. Consequently, transferring data obtained from these traffic control systems into a centralized database is essential for analytics purposes. Hence, this study aims to deploy a vehicle detection model and store the necessary data such as the vehicle class and count into a real-time database.