IDENTIFIKASI PLAT NOMOR KENDARAAN SECARA OFF-LINE BERBASIS PENGOLAHAN CITRA DAN JARINGAN SYARAF TIRUAN

Abstract

Vehicle number plate identification and recognition systems used in many applications, such as parking lot access, traffic control, and vehicles monitoring and security. In this research, an offline system to read and identify vehicle number plate is developed. Vehicle number plate identification process is performed in two steps, the first is the character image processing and the second is the pattern recognition with artificial neural networks. The character image processing consists of binerization, open-close morphology, median filtering, slicing and resizing. Back propagation algorithm is performed to train the neural networks,starting from the output layer to the input layer with output error is used to change network values. Testingis carried out in two scenarios, The first one is with the image samples which are included in the training step. The second one is with the image samples which are not included in the training step. The testing results show that the system is able to identify as much as 96 percent and 90 percent of number plate image samples for each scenarios.