Medical Waste Detection and Classification Through YOLO Algorithms

• Main Authors: Moktar M.H.B., Hajjaj S.S.H., Mohamed H.
• Other Authors: 57215719975
• Format: Conference paper
• Published: Springer Science and Business Media Deutschland GmbH 2025
• Subjects: Chemical wastes; Deep learning; Accidental injuries; Artificial intelligence methods; Classification models; Classifieds; Machine-learning; Medical wastes; Objects detection; Performances evaluation; Waste classification; Syringes

General waste is commonly managed to reduce pollution. Similarly, medical waste can be classified and managed to not only reduce pollution but also mitigate health risks and accidental injuries. Medical waste includes a variety of materials such as those contaminated with body fluids, sharps waste, and chemical waste. This study evaluates modern Artificial Intelligence methods for classifying medical waste such as facemasks, gloves, and syringes. Various classification models, including CNN, ResNet50, YOLO v3, and YOLO v4, were used and compared. YOLO v4 achieves a higher mAP (89.21%), surpassing YOLO v3 and other YOLO models used in waste classification studies. YOLO v4 was then tested in object detection and successfully identified masks, gloves, and syringes. Further performance evaluations are necessary to enhance the detection of medical waste and other objects in various applications. ? The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.