This paper presents a cost-effective and efficient deep learning-based system for the automated segregation of biodegradable (B) and non-biodegradable (NB) waste materials. The system integrates the YOLOv8 object detection algorithm with Python programming, Arduino Uno microcontroller, and hardware components such as a conveyor belt, IR sensor, relay, servo motor, and webcam. A custom dataset of biodegradable and non-biodegradable objects was collected, augmented, and used to train the YOLOv8 model. The trained model was integrated with Arduino to enable real-time object detection and sorting. When an object is detected on the conveyor belt, the system halts, captures an image, classifies the waste, and actuates the servo motor to direct the item into the appropriate bin. Experimental results demonstrate a mean Average Precision (mAP@0.5) of 98.25% and a sorting accuracy of 97.0%, significantly reducing manual effort and ensuring real-time reliability. This approach enhances recycling efficiency, reduces environmental impact, and contributes to sustainable waste management practices.
Mano. et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: