Real-Time Somali License Plate Recognition Using Deep Learning Model

The need for automatic license plate recognition is what is primarily driving the growing integration of computer technology in crucial industries like public transportation, healthcare parking, and retail parking. As cities grow, the interplay between technology and human needs becomes more obvious. In light of this trend, this paper presents a novel approach to license plate recognition in IoT-enabled smart parking systems, leveraging deep learning techniques. Traditional parking management systems often rely on manual monitoring or physical sensors, leading to inefficiencies and delays. In contrast, our proposed deep learning-based approach utilizes Convolutional Neural Networks (CNNs) for accurate license plate segmentation and character recognition. We curated a diverse dataset of Somali license plate images captured under various environmental conditions to train and evaluate our model. Through extensive experimentation, our model achieved an impressive accuracy rate of 96.76% after 80 epochs of training. Therefore, this research contributes to the advancement of efficient and accurate license plate recognition systems, facilitating enhanced parking management, traffic regulation, and urban mobility in smart cities.