Sistem Deteksi Kendaraan Overdimension secara Real-time Menggunakan Deep Learning di Jalan Raya

Nabih, Ahmad Hamayan (2024) Sistem Deteksi Kendaraan Overdimension secara Real-time Menggunakan Deep Learning di Jalan Raya. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5024201014-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (24MB) | Request a copy

Abstract

Kendaraan overdimension menjadi tantangan serius dalam lalu lintas di Indonesia. Mengakibatkan kerusakan infrastruktur jalan dan meningkatkan potensi kecelakaan, pendeteksian dan pengawasan terhadap pelanggaran overdimension menjadi penting. Penelitian ini mengembangkan sistem deteksi kendaraan overdimension di jalan raya secara realtime menggunakan metode deep learning, khususnya YOLO, dengan bantuan framework PyTorch. Data kendaraan dianotasi menggunakan Roboflow dan model dilatih menggunakan dataset yang relevan. Model kemudian diintegrasikan ke sistem kamera untuk deteksi realtime dan informasi dikirim ke dashboard web untuk pemantauan oleh pihak berwenang. Hasil awal menunjukkan efektivitas dan akurasi tinggi dalam mendeteksi pelanggaran overdimension, menjanjikan solusi yang lebih baik dalam mengawasi dan mengurangi pelanggaran lalu lintas.
======================================================================================================================================
Overdimension vehicles pose a significant challenge in Indonesian traffic. Leading to road infrastructure damage and increased accident potential, the detection and monitoring of overdimension violations become crucial. This study developed a real-time overdimension vehicle detection system on highways using deep learning, specifically the YOLO method, aided by the PyTorch framework. Vehicle data were annotated using Roboflow and the model was trained using a relevant dataset. The model was then integrated into a camera system for real-time detection, and the information was relayed to a web dashboard for monitoring by authorities. Preliminary results show high effectiveness and accuracy in detecting overdimension violations, promising a superior solution in overseeing and reducing traffic violations.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Dashboard Web, Deep Learning, Deteksi Realtime, Overdimension, PyTorch, Roboflow, YOLO, Deep Learning, Overdimension, PyTorch, Real-time Detection, Roboflow,Web Dashboard, YOLO.
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > TE Highway engineering. Roads and pavements > TE228.3 Intelligent transportation systems.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Ahmad Hamayan Nabih
Date Deposited:	16 Aug 2024 03:42
Last Modified:	16 Aug 2024 03:42
URI:	http://repository.its.ac.id/id/eprint/112824

Actions (login required)

View Item