Data Modeling and Integration for a Parking Management System with License Plate Recognition

This study demonstrates a method for integrating a YOLOR-based License Plate Recognition (LPR) module with a Firebase Realtime database (FRD) for Parking Management System (PMS) applications. A framework for implementing LPR technology within Philippine contexts is outlined, concentrating on the country's unique variations of four-wheeled motor vehicle license plates. Due to the distinctive features of license plate appearances, a custom image dataset must be used to train an LPR system's license plate detection model (LPD). The LPD model generates bounding box predictions to provide an Optical Character Recognition (OCR) region of interest (ROI). The extracted license plate text is then sent to the FRD. This database employs the NoSQL data modeling structure and is modeled with future possible database-dependent parking management system features considered. The contextual approach of the study is essential for addressing the unique demands of the Philippine context, providing possible insights and directions for similar initiatives across the country.