Vehicle Number Plate Recognition and Parking Authentication System

Abstract

Introduction: Introduction: Conventional manual or primitive automated systems employed for vehicle access control and parking management often suffer from inefficiencies, delays, and security risks. The accurate identification of vehicles at entry points, especially under varied real-world conditions, is a significant challenge that general-purpose solutions often fail to meet reliably, thus emphasizing the need for more sophisticated and reliable automated systems.

Objectives: The main objective of this project was to conceive, design, develop, and test a strong and smart system for automatic car access control and parking management. One of the main objectives was to offer security and efficiency in operation by developing a system that could supply accurate, real-time vehicle identification and authentication, thus specifically catering to the shortcomings involved in existing dependable license plate recognition methods.

Methods: A core computer vision pipeline based on an improved YOLOv8 architecture was used in the system. A Novel Parallel Attention Mechanism was introduced into YOLOv8, which supported multi-scale feature extraction through an Inception module specifically designed to improve the accuracy and reliability of vehicle license plate recognition. After the detection process, optical character recognition (OCR) was performed to enable accurate decoding of license plate information. The system also included a database to enable authentication processes and was interfaced with the Twilio API to enable real-time communication and notification. The functional capability and performance of the integrated system were validated through real-world tests.

Results: System testing showed that the system worked with excellent precision in detecting and reading vehicle license plates under a variety of conditions. Comparative testing showed that the specialized vision approach worked much better at the critical task of license plate detection than a standard approach. The system was also extremely fast and efficient in overall end-to-end operation, enabling real-time authentication.

Conclusions: We have successfully designed and demonstrated a robust automated Vehicle Access Control System. Through the use of an innovative attention-based vision system, combined with optical character recognition (OCR), database authentication, and real-time communication, we have developed an intelligent gatekeeper that is more secure and convenient. This project has established that the integrated solution successfully addresses the shortcomings of current systems, thereby offering a valid solution for contemporary access control and parking management requirements.