Penerapan Algoritma Backpropagation Pada Pengenalan Tanda Nomor Kendaraan Bermotor Untuk Kartu Parkir Berbasis RFID

Abstract

This research was conducted to improve a parking system's security using Radio Frequency Identification (RFID) technology to open parking barriers. To open the parking barriers, you must use an RFID card that has been registered to the system. There is no license plate checking on the parking system so that the RFID card can take other vehicles from the parking lot. Therefore, the RFID card needs to be linked with the license plate data using pattern recognition with an Artificial Neural Network. In this parking system, the backpropagation algorithm will be applied to identify the characters from the license plate's image captured with the camera. Information on parked vehicles can be seen through a web- based application that the administrator can only access. Also, this application can view the history of previously parked vehicles. Based on the calculation of accuracy using the confusion matrix, the backpropagation algorithm to identify the characters in this parking system has an accuracy value of 85.7%.