QuickTag: A Machine Learning-Based Automated Toll Collection System Using Optical Character Recognition

A gradual increase in the number of vehicles owned by the Indian population over the years has caused a lot of traffic at toll gates making it difficult to manually handle the gridlock, making the process cumbersome and time-consuming. Currently, toll collection is done using FASTag, which operates as a digital wallet. The toll is deducted when the operator scans the vehicle barcode, if the balance is insufficient the driver has to pay using cash which consumes time. It also results in the FASTag system being slow and cumbersome due to bad scans and no scans at a lot of places. The said problem was the genesis for developing an automated and improved system that can make our lives hassle-free and increase the efficiency of this operation to save time and energy. This automated system consists of OCR technology, i.e., optical character recognition that captures the license plate, and the owner is billed instantly using the auto-pay protocol. Vehicles pass through toll gates at a speed of 20–30 km/h. OCR-enabled speed cameras installed at the gates will capture the license plate, and text will be extracted from the image. In addition to that, a backup QR code will be installed on the windshield with vehicle and owner information encoded within it, QR code is to be used in case the license plate is muddy or not recognizable, this system will make the toll collection process more efficient without creating a traffic block at the toll plaza.