ITDS | SPMS

AUTOMATIC NUMBER PLATE RECOGNITION OF GHANAIAN VEHICLE USING YOLO

Project Description
Supervisor

The fast growing development in Ghana's major cities including Accra, Kumasi and Sunyani has led to an increase in vehicle ownership which has then reshaped the transportation systems. The said situation which has expanding economy through vehicle ownership has created new problems that include heavy traffic congestion and numerous traffic violations and stolen vehicles and fake license plates. The traditional manual vehicle identification system based on human observation and paper records has become insufficient because it produces errors and delays while struggling to handle the increasing number of vehicles. The current inefficient systems have created two major problems which endanger public safety and exhaust police resources while taking away their ability to handle essential security matters. The implementation of Automatic Number Plate Recognition technology provides a solution to Ghana's challenges because it uses computer vision to perform vehicle identification for various applications including traffic management and security enforcement. The implementation of such a system in Ghana, faces specific challenges because the country lacks standardized car license plates and uses different fonts, size and image quality varies and environmental conditions like dust and changing light affect recognition precision. This project aims to create an ANPR system tailored for Ghanaian needs which uses advanced AI to develop an efficient and reliable solution for vehicle identification. Our solution depends on YOLOv11 deep learning model which excels at object detection to achieve precise license plate detection. The system uses Tesseract OCR for character recognition to extract and read plate information from images regardless of their quality. We established our own dataset because there was no existing database which contained Ghanaian license plate images so we collected 1000 pictures of local vehicles which included private and commercial, government and special plates throughout Sunyani town, neighbouring areas and also Fiapre town. We the system developers selected images that demonstrated real-world variations to build a robust system which works with different plate designs and environmental conditions. The ANPR system uses Python, OpenCV and Ultralytics framework to perform advanced preprocessing operations which improve its performance. The system maintains high accuracy in plate detection and character extraction through these methods which handle image noise and low issues during the performance of activities with dust and inadequate light. The overall structure of formulating provides automatic identification operations and flexible possibility of application, i.e., traffic monitoring surveillance, toll operations, parking management and police assistance among others. Encouraging public safety by reducing human input in verification processes through our solution will free up resources for public initiatives as and also reduced errors and faster outputs. Overall, the research demonstrated how Artificial Intelligence can transform transport in Ghana with solutions to new urban mobility problems or cases. This automatic number plate recognition system was adjusted to work specifically to Ghanaian standards or specifications; it serves as a stepping stone for the usage of similar technology for the rest of the region.

Machine Learning: Machine Learning (ML) research in Computer Science and Information Technology focuses on the development of algorithms and models that enable computers to learn from data and improve their performance over time without being explicitly programmed. It is a subset of Artificial Intelligence that uses statistical techniques to give machines the ability to learn patterns, make decisions, and predict outcomes based on data. Supervised learning, a key area of ML research, involves training models on labeled data, where the input-output relationships are predefined. This method is widely used for tasks such as classification (e.g., spam detection) and regression (e.g., predicting house prices). Unsupervised learning, on the other hand, involves finding hidden patterns in data without predefined labels, with clustering and association being typical applications in areas such as customer segmentation and anomaly detection. Reinforcement learning is another area of ML that focuses on teaching agents to make decisions by interacting with their environment and receiving feedback in the form of rewards or penalties. It is often applied in robotics, game playing, and autonomous systems, where continuous learning and adaptation are required.