PERANCANGAN SISTEM DETEKSI TRAFFIC COUNTING MENGGUNAKAN YOLO VERSI 4

AZMINA ZATA ISMAH, . (2025) PERANCANGAN SISTEM DETEKSI TRAFFIC COUNTING MENGGUNAKAN YOLO VERSI 4. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text cover (2).pdf Download (762kB)
	Text bab 1.pdf Download (298kB)
	Text bab 2.pdf Restricted to Registered users only Download (642kB) | Request a copy
	Text bab 3.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text bab 4.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text bab 5.pdf Restricted to Registered users only Download (79kB) | Request a copy
	Text dafpus.pdf Download (218kB)
	Text lampiran.pdf Restricted to Registered users only Download (228kB) | Request a copy

Abstract

Revolusi industri keempat telah mendorong transformasi besar dalam berbagai aspek kehidupan, termasuk pada bidang transportasi dan teknologi informasi. Salah satu implikasinya adalah pemanfaatan kecerdasan buatan (AI) dalam sistem deteksi objek untuk mendukung pengambilan keputusan berbasis data, seperti dalam sistem traffic counting. Penelitian ini bertujuan untuk mengembangkan sistem deteksi dan penghitungan jenis kendaraan secara otomatis menggunakan algoritma YOLO (You Only Look Once) berbasis video, guna menggantikan metode manual yang memakan waktu dan biaya. Sistem ini dirancang untuk mengenali 8 kelas kendaraan sesuai klasifikasi Bina Marga dan diuji menggunakan YOLOv4. Berdasarkan hasil evaluasi, YOLOv4 memperoleh mAP 78%. Model yang telah dilatih kemudian diimplementasikan ke dalam situs web menggunakan framework Flask. Sistem ini diharapkan dapat membantu pihak terkait, seperti Dinas Perhubungan, dalam meningkatkan efisiensi pengumpulan data lalu lintas dan mendukung pengambilan kebijakan transportasi yang lebih tepat. Kata kunci: deteksi objek, YOLO, traffic counting, kendaraan. ***** The fourth industrial revolution has driven major transformations in various aspects of life, including transportation and information technology. One implication is the use of artificial intelligence (AI) in object detection systems to support data-driven decision-making, such as in traffic counting systems. This research aims to develop an automatic vehicle detection and counting system using the video-based YOLO (You Only Look Once) algorithm, to replace time-consuming and costly manual methods. This system is designed to recognize 8 vehicle classes according to the Bina Marga classification and was tested using YOLOv4. Based on the evaluation results, YOLOv4 achieved a mAP of 78%. The trained model was then implemented into a website using the Flask framework. This system is expected to assist relevant parties, such as the Transportation Agency, in increasing the efficiency of traffic data collection and supporting more appropriate transportation policy making. Keywords: object detection, YOLO, traffic counting, vehicle.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Irma Permata Sari, M. Eng. ; 2). Murien Nugrahaeni, M. Cs.
Subjects:	Teknologi dan Ilmu Terapan > Teknik Komputer
Divisions:	FT > S1 Sistem dan Teknologi Informasi
Depositing User:	Azmina Zata Ismah .
Date Deposited:	04 Aug 2025 06:48
Last Modified:	04 Aug 2025 06:48
URI:	http://repository.unj.ac.id/id/eprint/57672

Actions (login required)

View Item