PENGGUNAAN DEEP LEARNING UNTUK KLASIFIKASI KENDARAAN BERBASIS CITRA DALAM KAWASAN TERTIB LALU LINTAS DI KABUPATEN SUMEDANG

Rival Ramadhan, - (2022) PENGGUNAAN DEEP LEARNING UNTUK KLASIFIKASI KENDARAAN BERBASIS CITRA DALAM KAWASAN TERTIB LALU LINTAS DI KABUPATEN SUMEDANG. S1 thesis, Universitas Pendidikan Indonesia.

[img]
Preview
Text
S_ PSTI _1805181_Title.pdf

Download (741kB) | Preview
[img]
Preview
Text
S_ PSTI _1805181_Chapter1.pdf

Download (115kB) | Preview
[img] Text
S_ PSTI _1805181_Chapter2.pdf
Restricted to Staf Perpustakaan

Download (181kB)
[img]
Preview
Text
S_ PSTI _1805181_Chapter3.pdf

Download (550kB) | Preview
[img] Text
S_ PSTI _1805181_Chapter4.pdf
Restricted to Staf Perpustakaan

Download (1MB)
[img]
Preview
Text
S_ PSTI _1805181_Chapter5.pdf

Download (37kB) | Preview
[img] Text
S_ PSTI _1805181_Appendix.pdf
Restricted to Staf Perpustakaan

Download (2MB)
Official URL: http://repository.upi.edu/

Abstract

Penelitian ini dilatar belakangi oleh banyaknya pelanggar kawasan tertib lalu lintas. Dalam kawasan tertib lalu lintas di Kab. Sumedang, ada beberapa jenis kendaraan yang tidak diperbolehkan untuk memasuki kawasan tersebut. Adanya ATCS belum dimanfaatkan secara maksimal oleh Dinas Perhubungan terutama pada bidang deep learning. Penelitian ini bertujuan untuk melakukan penerapan program pengolahan citra untuk mendeteksi kendaraan menggunakan algoritma YOLOv5 melalui rekaman CCTV lalu lintas; melakukan pengujian terhadap hasil pengolahan citra kendaraan melalui rekaman CCTV lalu lintas; dan memberikan rekomendasi pengembangan sistem pendeteksian kendaraan pada kawasan tertib lalu lintas di Kab. Sumedang. Penelitian ini menerapkan AI Project Cycle dan algoritma YOLOv5 sebagai algoritma yang digunakan untuk mendeteksi kendaraan. Dataset didapat melalui video yang ada pada website ATCS Dinas Perhubungan Kab. Sumedang. Berdasarkan hasil penelitian diperoleh kesimpulan bahwa arsitektur yang digunakan melalui hasil dari training serta validasi dalam model YOLOv5 berhasil mendeteksi dengan akurat; serta model ini dapat mendeteksi keempat jenis kendaraan dengan cukup baik dengan mendapatkan nilai mAP di semua kelas sebesar 91,6%. -----The background of this research is based on the number of violators in the traffic order area. In the traffic order area in Sumedang Regency, several types of vehicles are not allowed to enter the area. The existence of ATCS has not been fully utilized by the Department of Transportation, especially in the field of deep learning. This study aims to implement an image processing program to detect vehicles using the YOLOv5 algorithm through traffic CCTV recordings; conduct testing on the results of vehicle image processing through CCTV traffic recordings; and provide architectural recommendations for the development of vehicle detection systems in traffic orderly areas in Kab. Sumedang. This study applies the AI Project Cycle and uses YOLOv5 as the algorithm that will be used to detect vehicles. The dataset used was obtained through a video on the ATCS website of the Department of Transportation in Kab. Sumedang. Based on the results of the study, it was concluded that the architecture used through the results of training and validation in the YOLOv5 model managed to detect accurately; and this model can detect the four types of vehicles quite well by getting an average mAP value in all classes of 91.6%.

Item Type: Thesis (S1)
Additional Information: Link Google Scholar: https://scholar.google.com/citations?hl=id&authuser=1&user=OBXz7YEAAAAJ ID Sinta Dosen Pembimbing: Ahmad Fauzi : 6122861 Rian Andrian : 6681695
Uncontrolled Keywords: Deep learning, Deteksi Kendaraan, AI Project Cycle, YOLOv5
Subjects: L Education > L Education (General)
T Technology > T Technology (General)
Divisions: UPI Kampus Purwakarta > S1 Pendidikan Sistem Teknologi dan Informasi
Depositing User: Rival Ramadhan
Date Deposited: 07 Sep 2022 00:48
Last Modified: 07 Sep 2022 00:48
URI: http://repository.upi.edu/id/eprint/78284

Actions (login required)

View Item View Item