Real-Time Vehicle Detection and Air Pollution Estimation Using YOLOv9
Pollution of air, particularly in cities, is becoming an issue to be taken seriously owing to the health and environmental risks associated with it, and the major contributor to air pollution is car emissions. The objective of the study is to identify and classify vehicles such as motorbikes, cars,...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
LPPM ISB Atma Luhur
2025-01-01
|
| Series: | Jurnal Sisfokom |
| Subjects: | |
| Online Access: | https://jurnal.atmaluhur.ac.id/index.php/sisfokom/article/view/2339 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823856510328897536 |
|---|---|
| author | Hari Suparwito Bernardus Galih Hersa Prakoso Rosalia Arum Kumalasanti Agnes Maria Polina |
| author_facet | Hari Suparwito Bernardus Galih Hersa Prakoso Rosalia Arum Kumalasanti Agnes Maria Polina |
| author_sort | Hari Suparwito |
| collection | DOAJ |
| description | Pollution of air, particularly in cities, is becoming an issue to be taken seriously owing to the health and environmental risks associated with it, and the major contributor to air pollution is car emissions. The objective of the study is to identify and classify vehicles such as motorbikes, cars, buses, trucks in order to monitor live traffic and potentially determine the extent to which the pollution level elevates, utilizing the YOLOv9 model. Traffic CCTV camera footage was gathered under a wide range of circumstances including different lighting and varying traffic intensity. Folders were particularly structured and images annotated, in the manner, which served the purpose of meeting the requirements of the YOLO structure. Once it was trained with a labeled dataset, the vehicle identification by YOLOv9 model was found to be quite satisfactory. Overall vehicle identification accuracy was calculated to be mAP50:95 of 0.826. In contrast, it had a harder time with smaller items like motorcycles, with a mAP50:95 of 0.682. Findings indicate that larger items were detected more than smaller items. Camera angles and the small size of the objects often make small objects appear to blend in to the background. This research indicates that AI can be of help when dealing with the urban structure. It offers a way of measuring traffic volume to predict the amount of CO emissions that can be avoided or controlled. The rest are keen in enhancing the effectiveness of recognizing small objects within the system and deploying it in multiple settings. |
| format | Article |
| id | doaj-art-8704a4946cfd45e7a24576b055c43033 |
| institution | Kabale University |
| issn | 2301-7988 2581-0588 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | LPPM ISB Atma Luhur |
| record_format | Article |
| series | Jurnal Sisfokom |
| spelling | doaj-art-8704a4946cfd45e7a24576b055c430332025-02-12T07:27:38ZengLPPM ISB Atma LuhurJurnal Sisfokom2301-79882581-05882025-01-01141233010.32736/sisfokom.v14i1.23392002Real-Time Vehicle Detection and Air Pollution Estimation Using YOLOv9Hari Suparwito0Bernardus Galih Hersa Prakoso1Rosalia Arum Kumalasanti2Agnes Maria Polina3Department of Informatics, Universitas Sanata DharmaDepartment of Informatics, Universitas Sanata DharmaDepartment of Informatics, Universitas Sanata DharmaDepartment of Informatics, Universitas Sanata DharmaPollution of air, particularly in cities, is becoming an issue to be taken seriously owing to the health and environmental risks associated with it, and the major contributor to air pollution is car emissions. The objective of the study is to identify and classify vehicles such as motorbikes, cars, buses, trucks in order to monitor live traffic and potentially determine the extent to which the pollution level elevates, utilizing the YOLOv9 model. Traffic CCTV camera footage was gathered under a wide range of circumstances including different lighting and varying traffic intensity. Folders were particularly structured and images annotated, in the manner, which served the purpose of meeting the requirements of the YOLO structure. Once it was trained with a labeled dataset, the vehicle identification by YOLOv9 model was found to be quite satisfactory. Overall vehicle identification accuracy was calculated to be mAP50:95 of 0.826. In contrast, it had a harder time with smaller items like motorcycles, with a mAP50:95 of 0.682. Findings indicate that larger items were detected more than smaller items. Camera angles and the small size of the objects often make small objects appear to blend in to the background. This research indicates that AI can be of help when dealing with the urban structure. It offers a way of measuring traffic volume to predict the amount of CO emissions that can be avoided or controlled. The rest are keen in enhancing the effectiveness of recognizing small objects within the system and deploying it in multiple settings.https://jurnal.atmaluhur.ac.id/index.php/sisfokom/article/view/2339air pollutioncarbon monoxide computer visionobject detectionyolov9 |
| spellingShingle | Hari Suparwito Bernardus Galih Hersa Prakoso Rosalia Arum Kumalasanti Agnes Maria Polina Real-Time Vehicle Detection and Air Pollution Estimation Using YOLOv9 Jurnal Sisfokom air pollution carbon monoxide computer vision object detection yolov9 |
| title | Real-Time Vehicle Detection and Air Pollution Estimation Using YOLOv9 |
| title_full | Real-Time Vehicle Detection and Air Pollution Estimation Using YOLOv9 |
| title_fullStr | Real-Time Vehicle Detection and Air Pollution Estimation Using YOLOv9 |
| title_full_unstemmed | Real-Time Vehicle Detection and Air Pollution Estimation Using YOLOv9 |
| title_short | Real-Time Vehicle Detection and Air Pollution Estimation Using YOLOv9 |
| title_sort | real time vehicle detection and air pollution estimation using yolov9 |
| topic | air pollution carbon monoxide computer vision object detection yolov9 |
| url | https://jurnal.atmaluhur.ac.id/index.php/sisfokom/article/view/2339 |
| work_keys_str_mv | AT harisuparwito realtimevehicledetectionandairpollutionestimationusingyolov9 AT bernardusgalihhersaprakoso realtimevehicledetectionandairpollutionestimationusingyolov9 AT rosaliaarumkumalasanti realtimevehicledetectionandairpollutionestimationusingyolov9 AT agnesmariapolina realtimevehicledetectionandairpollutionestimationusingyolov9 |