Virtual Booth

The process of recognizing and sorting the recycling material is a critical aspect in waste management. Presently, there are many waste management companies who still use the manual classification approach. This approach is tedious and time-consuming. Additionally, manual classification approach requires high resources that may lead to the increase of operational cost. This can be inferred that an automated recycling material platform is urgently needed. Therefore, this project presents a system namely, Recycle Go! : A Real-Time Recycling Material Recognition System. This system focuses on recognizing three type of recycling materials which are plastic bottles, aluminum cans and paper. The well-known YOLOv8 algorithm was applied to train the object detection and recognition model. The accuracy of the trained model was assessed using metrics like confusion matrix, recall, average precision, and F1 score. Based on the achieved results, the mean average precision for all classes surpasses 90.00%, and the F1 score for all classes exceeds 0.9 accuracy. Thus, this can be concluded that the developed Recycle Go! system is able to perform good recycling material recognition task. It is believed that this Recycle Go! can be further expand by enhancing the algorithm and to cover more types of recycling materials.