Comparison of the Vehicle’s License Plate Recognition Using Image Centroid and Zone with K-Nearest Neighbour and Probabilistic Neural Network

Nama Peneliti/Mahasiswa HAFARA FISCA LAHMURAHMA (G64096027) Judul Perbandingan dalam Pengenalan Karakter Plat Nomer Kendaraan Menggunakan Image Centroid And Zone dengan Klasifikasi k-Nearest Neighbour dan Probabilistic Neural Network Judul ( English ) Comparison of the Vehicle’s License Plate Recognition Using Image Centroid and Zone with K-Nearest Neighbour and Probabilistic Neural Network Pembimbing/Supervisor Mushthofa Abstrak/Ringkasan HAFARA FISCA LAHMURAHMA. Comparison of the Vehicle’s License Plate Recognition Using Image Centroid and Zone with K-Nearest Neighbour and Probabilistic Neural Network. Supervised by MUSHTHOFA SKom, MSc. Pattern recognition system is a scientific concept which is being developed with a variety of methods. One example is the vehicle license plate character recognition. License plate has a serial number, that is, the arrangement of letters and numbers that are specific to the vehicle. This study uses the method of image centroid and zone (ICZ) feature extraction with 14 zones. In addition, this study uses two classification methods, namely k-nearest neighbour (KNN) and probabilistic neural network (PNN), and aims to compare the accuracy of both methods.  Classification are done for 100 vehicle license plate images, as the data. Values of k in KNN are 1, 3, 5, 7, and 9, while the bias values used in the PNN are 0.1, 0.2, 0.8, 1, 2, 4, and 8. The highest accuracy for vehicle’s license plate recognition using KNN classification method is 60.00%, whereas the highest accuracy for PNN classification method is 58.46%. Keywords : image centroid and zone (ICZ), k-nearest neighbour (KNN), license plate, probabilistic neural network (PNN)