Desain Sistem Pendeteksi Plat Kendaraan Untuk Pengisian BBM Menggunakan Metode K-Nearest Neighbor (KNN)

Abstract

This study aims to design and develop a vehicle plate detection system for refueling. Applications built using the K-Nearest Neighbor (KNN) method, this method is used to classify vehicle plates. This research will use the OpenCV module in the python programming language to recognize vehicle plates and carry out the pre-processing, segmentation, feature extraction, and classification stages of vehicle Image data. This system will utilize the K-Nearest Neighbor (KNN) method such as edge detection, perspective transformation, and pattern recognition to achieve optimal results. The test results show that the system is able to recognize with high accuracy various types of vehicle license plates under different lighting conditions and viewing angles, thus a system can be created and determined for each license plate to only be refueled once a day. When the vehicle has never been filled on that day, the system will display a green light which means refueling can be done. Whereas for vehicles that have filled in on that day, the system will display a red light which means the vehicle cannot fill up so it must wait the next day after the system is reset. This system will be able to accurately and real-time detect and identify the vehicle plates used in the refueling process. Therefore, the design of this system is expected to be the basis for increasing efficiency in the refueling industry as well as providing a basis for the development of further research in this field.