A license plate detection and recognition method using YOLOv5n and LPRNet / Wan Xing, Juliana Johari and Fazlina Ahmat Ruslan

Efficient and speedy license plate identification is crucial in vehicle-to-vehicle (V2V) communication scenarios for applications like traffic management and law enforcement. This study introduces a method that utilizes cutting-edge deep learning models to identify and recognize license plates on vehicles. The YOLOv5n model, renowned for its exceptional detection precision and efficient design, is utilized for license plate detection. After the detection phase, the License Plate Recognition Network (LRPNet) is employed to precisely identify the characters on the detected license plates. The thorough assessment we conducted evaluates the performance of this integrated system on public datasets, showcasing its resilience and effectiveness. This paper provides an in-depth discussion of the license plate detection and recognition task through very detailed theoretical derivations and structural block diagrams combined with experiments. The findings indicate that the utilization of YOLOv5n and LRPNet shows the high precision and efficiency of license plate detection and recognition.