부산대학교 도서관

Number Plate Recognition using image processing and Optical Character Recognition (OCR) techniques aims to create an efficient automatic system for authorized vehicle identification. The proposed method involves a comprehensive iterative waterfall model for software development, emphasizing flexibility and risk management Challenges include handling diverse fonts, languages, and image variations, such as dear, blurred, and skewed plates. The system employs OpenCV for image manipulation, Tesseract OCR for text extraction, and Pandas for database matching. The objectives encompass accurate license plate detection, character recognition, and extraction of vehicle information from a dataset. The model demonstrates an 83% accuracy rate and offers potential applications in traffic management, surveillance, and toll collection. Future enhancements include extending the system to detect international plates, various languages, real-time applications, and integration into traffic infrastructure for enhanced functionality.