International Journal of Science, Engineering and TechnologyAuthors: Dinesh V N, Dharani G Assistant Professor, Santhosh N, Nagadharshini M
Abstract: Parking management systems are advancing with the integration of Computer Vision and Automatic Number Plate Recognition technologies to enhance security, efficiency, and automation. This project aims to create a smart parking system that autonomously verifies vehicle registrations and allocates parking spaces to authorized vehicles while denying access to unregistered ones. ANPR technology plays a key role in vehicle identification, ensuring secure entry. However, existing systems struggle to accurately detect plates under conditions like poor lighting, bad weather, or damaged plates, which can impact reliability, especially in high-traffic areas.To address these issues, the project uses advanced image processing techniques such as adaptive thresholding, image denoising, and modern Optical Character Recognition (OCR) models like EasyOCR and PaddleOCR. These models improve plate detection under unclear images, varied fonts, and inconsistent lighting. Additionally, the system employs GPU acceleration, multithreading, and optimized database queries to reduce processing delays, ensuring real-time operation even in crowded environments.System performance will be assessed based on Security,Efficiency,Parking Management Automation,License Plate Detection & Recognition Performance.The system's ability to detect number plates accurately under different conditions, process vehicle entry requests efficiently, and allocate parking spaces while preventing unauthorized access will be key factors in its success. This approach aims to deliver a secure, efficient, and automated parking management solution for high-traffic areas.
