Sistem Pengenalan Pelat Nomor Kendaraan Untuk Mendukung Kebijakan Ganjil-Genap

Wahyutama, I Putu Adhista (2023) Sistem Pengenalan Pelat Nomor Kendaraan Untuk Mendukung Kebijakan Ganjil-Genap. Other thesis, Institut Teknonogi Sepuluh Nopember.

[thumbnail of 5022201246-Undergraduate_Thesis.pdf] Text
5022201246-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 October 2026.

Download (9MB) | Request a copy

Abstract

Berbagai usaha untuk mengurangi kepadatan lalu lintas telah dilakukan, baik oleh pihak kepolisian maupun dinas perhubungan, salah satu diantaranya adalah penerapan sistem ganjil-genap. Dalam implementasinya sistem ganjil-genap tersebut terkadang tidak efektif karena tidak ada pihak yang senantiasa memantau terjadinya pelanggaran, kecuali hanya sesekali jika ada petugas yang sedang melakukan penertiban. Padahal dampak dari penerapan sistem ganjil genap tidak hanya membantu kelancaran lalu lintas namun juga berperan dalam mengurangi polusi udara. Berangkat dari permasalahan tersebut dalam tugas akhir ini, mengajukan gagasan inovatif dalam bentuk aplikasi sistem tilang ganjil genap yang diberi nama Sistem Tilang Elektronik Ganjil Genap Otomatis (STEGGO). Sistem STEGGO dikembangkan berbasis Artificial Inteligent (AI) dengan menggunakan deep learning metode Convolutional Neural Network (CNN) dan melibatkan algoritma You Look Only Once (YOLO) untuk menyelesaikan tugas deteksi lokasi pelat nomor dan library EasyOCR untuk pekerjaan segmentasi karakter dan pengenalan karakter. Untuk menguji keberhasilan sistem yang dibuat, sistem STEGGO telah diujicobakan pada 50 citra mobil, 50 citra sepeda motor, dan 45 citra video yang berisi gabungan mobil dan sepeda motor. Dari berbagai skenario pengujian yang telah dilakukan, sistem STEGGO memiliki tingkat keberhasilan sebesar 100% dalam hal deteksi lokasi pelat nomor, 98.7% dalam proses segmentasi karakter dan 96.33% pada tahap pengenalan karakter. Pada pengujian sistem ganjil-genap sistem STEGGO juga telah berhasil mengklasifikasikan pelat nomor ganjil-genap secara baik dengan tingkat kesalahan hanya 2.07%. Sementara untuk pengujian status tilang performansi sistem STEGGO bisa mencapai 97.77%. Dengan mempertimbangkan berbagai skenario pengujian yang ada maka dapat disimpulkan bahwa sistem STEGGO mempunyai tingkat kinerja sebesar 98,06%.
=================================================================================================================================
Various efforts to reduce traffic congestion have been made, both by the police and the transportation department, one of which is the implementation of the odd-even system. In its implementation, the odd-even system is sometimes ineffective because there are no parties who always monitor the occurrence of violations, except only occasionally if there are officers who are conducting enforcement. Whereas the impact of the implementation of the even-odd system not only helps smooth traffic but also plays a role in reducing air pollution. Departing from these problems, this final project proposes an innovative idea in the form of an odd-even ticketing system application called the Automatic Odd-Even Electronic Ticket System (STEGGO). The STEGGO system is developed based on Artificial Intelligence (AI) by using deep learning Convolutional Neural Network (CNN) method and involving You Look Only Once (YOLO) algorithm to complete the license plate location detection task and EasyOCR library for character segmentation and character recognition work. To test the success of the system, the STEGGO system has been tested on 50 car images, 50 motorcycle images, and 45 video images containing cars and motorcycles combined. From the various test scenarios that have been carried out, the STEGGO system has a success rate of 100% in terms of license plate location detection, 98.7% in the character segmentation process and 96.3% in the character recognition stage. In testing the odd-even system, the STEGGO system has also succeeded in classifying odd-even license plates well with an error rate of only 2.07%. Meanwhile, for testing the ticket status, the performance of the STEGGO system can reach 97.77%. By considering various test scenarios, it can be concluded that the STEGGO system has a performance level of 98.06%.

Item Type: Thesis (Other)
Uncontrolled Keywords: Deep learning, EasyOCR, Ganjil-Genap, Pelat Nomor, license plate, odd-even, YOLO.
Subjects: T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
T Technology > TA Engineering (General). Civil engineering (General) > TA1650 Face recognition. Optical pattern recognition.
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5101 Telecommunication
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User: I Putu Adhista Wahyutama
Date Deposited: 29 Jul 2024 17:40
Last Modified: 30 Jul 2024 01:08
URI: http://repository.its.ac.id/id/eprint/109830

Actions (login required)

View Item View Item