Enhanced Gate Security System using CNN for Iraqi Vehicle and License Plate Detection

Abstract

 This paper presents the development of an advanced gate security system using Convolutional Neural Networks (CNNs) for the detection and recognition of Iraqi vehicles and their license plates. The proposed architecture consists of an image acquisition module, a preprocessing and segmentation pipeline, a CNN-based license plate recognition model, and a gate control unit integrated with a stepper motor and Arduino controller. The system works on a dataset of Iraqi license plate images obtained from public sources, standing for various lighting conditions, angles, and font styles to ensure robustness. Experimental results show that the proposed model achieves perfect accuracy, with 100% precision, recall, and F1-score across all tested digit classes. These results show the system’s capability to accurately classify license plates, even under challenging scenarios such as poor lighting or partial occlusion. The high reliability and automation potential of this approach make it highly suitable for real-world gate control applications, especially in sensitive environments such as government buildings and secure facilities. This work highlights the importance of integrating AI-based solutions into modern security infrastructures to enhance access control with minimal human intervention and improved operational efficiency.