Real-Time Waste Classification Using YOLOv8n: A Deep Learning Approach

Year 2026, Volume: 2 Issue: 1, 1 - 16, 30.01.2026

Ibrahim M.a. Samak , Gökalp Tulum

https://doi.org/10.26650/d3ai.1734221

https://izlik.org/JA55KR82WM

Abstract

This study presents an artificial intelligence-based approach to real-time waste classification using a lightweight version of the YOLOv8 object detection model. The environmental impact of improper waste management is significant, with increasing volumes of waste in urban areas posing serious challenges. Through the use of deep learning and computer vision, this research aims to enable efficient, automatic classification of solid waste into ten distinct waste classes. A labeled image dataset was curated and preprocessed using Roboflow [8] and various data augmentation strategies. The YOLOv8n model was trained and evaluated using precision, recall, F1 score, and mean Average Precision (mAP) metrics. Results demonstrate that the proposed system achieves high accuracy and real-time performance, making it suitable for deployment in smart bins, mobile waste collection systems, and urban surveillance applications. These outcomes include a mean Average Precision at IoU 0.5 (mAP@0.5) of 0.98, precision of 0.95, and recall of 0.98. The study contributes to the growing body of AI applications in environmental sustainability by offering a scalable and practical solution to address solid waste classification challenges in real-world settings.

Keywords

Artificial Intelligence , Real-time detection , Deep Learning , Waste Garbage , YOLOv8n

References

• Turkish Statistical Institute (TÜİK). 2025. Greenhouse Gas Emissions Statistics, 1990–2023. Retrieved April 10, 2025 from https://data.tuik.gov.tr/Bulten/Index?p=Greenhouse-Gas-Emissions-Statistics-1990-2023-53974 google scholar
• Joseph Redmon, Santosh Divvala, Ross Girshick, and Ali Farhadi. 2016. You Only Look Once: Unified, Real-Time Object Detection. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. google scholar
• Ultralytics. 2025. Ultralytics Documentation. Retrieved March 19, 2025, https://docs.ultralytics.com google scholar
• Chien-Yao Wang, Alexey Bochkovskiy, and Hong-Yuan Mark Liao. 2022. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. arXiv:2207.02696. Retrieved April 30, 2025, http://arxiv.org/abs/2207.02696 google scholar
• Kaiming He, Georgia Gkioxari, Piotr Dollár, and Ross Girshick. 2017. Mask R-CNN. In Proceedings of the IEEE International Conference on Computer Vision. google scholar
• Muhammad Hassan, Raza Ali, and Vinod Kumar. 2022. Real-Time Instance Segmentation in Solid Waste Management. Journal of Environmental Informatics. google scholar
• Juan Terven and Diego Cordova-Esparza. 2023. A Comprehensive Review of YOLO Architectures in Computer Vision: From YOLOv1 to YOLOv8 and YOLO-NAS. Machines 5, 4 (2023). https://doi.org/10.3390/make5040083 google scholar
• YOLOv7 Trash Detection Group. 2023. YOLOv7 Trash Dataset v5 (11.12.2024). Roboflow Universe. Retrieved September 30, 2024 from https://universe.oboflow.com/yolov7-trash-detection-group/yolov7-trash-dataset-v5-05-0 google scholar