Deteksi Kendaraan Dengan Asap Berlebih Menggunakan Metode You Only Look Once (YOLO) Dan Intersection Over Union (IOU) - Submit Jurnal

Nasarany, Darrell Timothy (2025) Deteksi Kendaraan Dengan Asap Berlebih Menggunakan Metode You Only Look Once (YOLO) Dan Intersection Over Union (IOU) - Submit Jurnal. Bachelor thesis, Institut Teknologi Kalimantan.

	Text 11211029_cover.pdf Restricted to Registered users only until 4 October 2027. Download (356kB) | Request a copy
	Text 11211029_statement_of_authenticity.pdf Restricted to Registered users only until 4 October 2027. Download (240kB) | Request a copy
	Text 11211029_publishing_agreement.pdf Restricted to Registered users only until 4 October 2027. Download (268kB) | Request a copy
	Text 11211029_approval_sheet.pdf Restricted to Registered users only until 4 October 2027. Download (242kB) | Request a copy
	Text 11211029_preface.pdf Restricted to Registered users only until 4 October 2027. Download (95kB) | Request a copy
	Text 11211029_abstract_id.pdf Restricted to Registered users only until 4 October 2027. Download (48kB) | Request a copy
	Text 11211029_abstract_en.pdf Restricted to Repository staff only until 4 October 2027. Download (47kB) | Request a copy
	Text 11211029_table_of_content.pdf Restricted to Repository staff only until 4 October 2027. Download (78kB) | Request a copy
	Text 11211029_illustrations.pdf Restricted to Repository staff only until 4 October 2027. Download (76kB) | Request a copy
	Text 11211029_tables.pdf Restricted to Repository staff only until 4 October 2027. Download (66kB) | Request a copy
	Text 11211029_chapter_1.pdf Restricted to Repository staff only until 4 October 2027. Download (181kB) | Request a copy
	Text 11211029_chapter_2.pdf Restricted to Repository staff only until 4 October 2027. Download (853kB) | Request a copy
	Text 11211029_chapter_3.pdf Restricted to Repository staff only until 4 October 2027. Download (377kB) | Request a copy
	Text 11211029_chapter_4.pdf Restricted to Repository staff only until 4 October 2027. Download (1MB) | Request a copy
	Text 11211029_conclusions.pdf Restricted to Repository staff only until 4 October 2027. Download (117kB) | Request a copy
	Text 11211029_bibliography.pdf Restricted to Registered users only until 4 October 2027. Download (160kB) | Request a copy
	Text 11211029_enclosure.pdf Restricted to Repository staff only until 4 October 2027. Download (60kB) | Request a copy
	Text 11211029_paper.pdf Restricted to Repository staff only until 4 October 2027. Download (4MB) | Request a copy
	Text 11211029_Form. TA-020.pdf Restricted to Repository staff only until 4 October 2027. Download (72kB) | Request a copy
	Text 11211029_presentation.pdf Restricted to Repository staff only until 4 October 2027. Download (1MB) | Request a copy

Abstract

Berkaitan dengan pencemaran udara yang menjadi hal mendesak di dunia, terutama Emisi kendaraan bermotor menjadi salah satu penyebab utama pencemaran udara yang berkontribusi secara signifikan terhadap masalah kesehatan masyarakat dan kerusakan lingkungan. Penelitian ini menggunakan deep learning terkhususnya Convolutional Neural Network (CNN). Metode yang digunakan yaitu You only Look once (YOLO) dan implementasi Intersection over Union (IoU) untuk menemukan serta menentukan kendaraan yang memiliki polutan kendaraan melalui deteksi asap dan kendaraan. Dataset pada penelitian berasal dari Roboflow yang dikumpulkan dengan pembagian data pelatihan (80%), data validasi (10%), dan data pengujian (10%). Pelatihan 4 model variasi YOLOv8 menghasilkan variasi pada tiap model, YOLOv8 Nano menghasilkan precision sebesar 85.5%, recall sebesar 91.6%, mAP sebesar 88.5% dan F1-Score sebesar 88.4%. Pada model YOLOv8 Small menghasilkan precision sebesar 86.8%, recall sebesar 89.7%, mAP sebesar 88.2% dan F1-Score sebesar 88.2%. Pada model YOLOv8 Medium menghasilkan precision sebesar 88.1%, recall sebesar 88.7%, mAP sebesar 88.4% dan F1-Score sebesar 88.4%. Pada model YOLOv8 Large menghasilkan precision sebesar 87.8%, recall sebesar 88.9%, mAP sebesar 88.4% dan F1-Score sebesar 88.3%. Setiap model mampu mendeteksi asap polusi kendaraan dengan baik. Implementasi IoU juga dapat membuktikan penentuan kepemilikan asap kendaraan berdasarkan irisan antara prediksi bounding box asap dan kendaraan. Berdasarkan hasil yang didapat diharapkan dapat menjadi landasan untuk mengembangkan sistem yang berkontribusi terhadap pencemaran udara, serta memberikan solusi yang lebih efektif dalam memonitor dan menanggulangi masalah ini.

Item Type:	Thesis (Bachelor)
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science T Technology > TL Motor vehicles. Aeronautics. Astronautics
Divisions:	Jurusan Matematika dan Teknologi Informasi > Informatika
Depositing User:	Darrell Timothy Nasarany
Date Deposited:	10 Jul 2025 02:59
Last Modified:	10 Jul 2025 02:59
URI:	http://repository.itk.ac.id/id/eprint/22948

Actions (login required)

View Item