<p>Efficient vehicle monitoring and parking management are essential for modern smart city infrastructure. This study presents a real-time vehicle number plate recognition and automated parking allocation system using deep learning and OCR techniques. The framework employs YOLOv8 for fast and accurate license plate detection in live video streams, followed by text extraction with EasyOCR and Tesseract. A post-processing module refines the extracted text by correcting misclassified characters and validating formats based on country-specific rules. For operational efficiency, a rule-based parking allocation assigns vehicles to East Wing (first 10 slots) or West Wing (next 10 slots), triggering a real-time “Parking Full” notification when all 20 spaces are occupied. Vehicle details, including number, timestamp, and slot assignment, are logged into Excel sheets for monitoring. The system was evaluated on a custom real-world dataset of over 5,000 annotated vehicle images. Experimental results show strong performance, achieving 98.5% detection accuracy, 98.2% precision, 97.8% recall, and 98.0% F1-score, with an average inference time of 50 ms per frame. Comparative analysis demonstrates improvements over existing approaches in recognition accuracy, speed, and deployability. The proposed solution provides a scalable, cost-effective framework suitable for commercial, residential, and institutional parking environments.</p>