Car License Plate Detection and Recognition Using Modified U-Net Deep Learning Model

Abstract

In ultramodern smart metropolises, licence plate recognition systems, similar to risk payment systems, parking risk payment systems, and apartment admission operation, are generally enforced. Similar electronic technology not only improves people's daily lives, but also provides governments with safe and effective services. People are clamouring for a safe life and trip in an era of rapid technological development. Over the last several decades, the no. of buses on the road has significantly increased. Tracking every vehicle can be a difficult process due to the tremendous diurnal development of vehicle assiduity. In this investigation, we propose an automatic vehicle monitoring system for fast-moving vehicles that makes use of a roadside surveillance camera. As a result, an integrated deep neural network has been proposed, capable of locating the licence plate and recognising characters in a single forward pass. Unlike previous systems, which treat licence plate recognition and identification as separate problems to be solved separately, our method employs a single network to achieve both goals at the same time. It not only eliminates intermediate errors, but it also speeds up processing. Our method is to convert the video to a photograph and then identify the car at each frame. The licence plate from the identified vehicle is then read. Finally, the read licence plate is used to commemorate the characters on the honoured licence plate. Fictitious deep literacy models use the Image AI library to greatly used for training purpose.