IMPLEMENTASI SUPPORT VECTOR MACHINE UNTUK PENGENALAN PLAT NOMOR KENDARAAN
Exact Papers in Compilation (EPiC) · 2019
DOI: 10.32764/epic.v1i3.156 ↗Kurz gefasst
Fachartikel in der Zeitschrift Exact Papers in Compilation (EPiC), 2019. Verfasst von Tholib Hariono.
Abstract (Originalsprache)
License plate recognition recognition has a very important role in developing better transport systems such as electronic tolls, electronic parking, traffic monitoring activities and others. Number plate recognition process is carried out through four main stages (plate detection, segmentation, feature extraction and classification). Plate detection is done to obtain the location of license plate using the Viola Jones algorithm. The segmentation process to separate the characters from the image of the license plate area-based color using labeling techniques. Feature extraction is based on the division of the image into several blocks and then calculates the average value of pixels every block using the integral image method to obtain characteristics. In classification using SVM with 3 kernels: linear, RBF and polynomials. The test results character acquired 93.75% accuracy is best in the division images of 50 blocks with polynomial kernel, and 82.86% accuracy in testing plate.
Zitieren
APA
Hariono, T. (2019). IMPLEMENTASI SUPPORT VECTOR MACHINE UNTUK PENGENALAN PLAT NOMOR KENDARAAN. Exact Papers in Compilation (EPiC). https://doi.org/10.32764/epic.v1i3.156
Vancouver
Hariono T. IMPLEMENTASI SUPPORT VECTOR MACHINE UNTUK PENGENALAN PLAT NOMOR KENDARAAN. Exact Papers in Compilation (EPiC). 2019. doi:10.32764/epic.v1i3.156
BibTeX
@article{Hariono2019,
author = {Hariono, Tholib},
title = {IMPLEMENTASI SUPPORT VECTOR MACHINE UNTUK PENGENALAN PLAT NOMOR KENDARAAN},
journal = {Exact Papers in Compilation (EPiC)},
year = {2019},
doi = {10.32764/epic.v1i3.156}
}
EUVIMED