Detection and Identification of Moroccan License Plates Using Convolutional Neural Networks

Abstract

License plate recognition using artificial intelligence has become a prominent research topic due to its crucial role in various applications, including road safety, smart traffic monitoring, parking management, toll collection, access control, and vehicle identification in law enforcement. In this study, we propose a practical method for detecting and recognizing Moroccan license plates, addressing challenges such as plate format, font style, and variability. Our approach employs two complementary convolutional neural network (CNN) models: one for recognizing Arabic letters and another for digits. This dual-model strategy improves overall recognition accuracy, and the models were trained on a large, diverse dataset of representative Moroccan license plates to ensure robustness and reliability.