Enhanced Plastic Detection and Classification: Advancing Recognition of Plastic Varieties Using YOLOv8

The widespread use of plastics in everyday life has caused serious problems related to the management of plastic waste. Problems related to knowledge about the types of plastics, namely PET/PETE (Polyetylene Terephthalate), HDPE (Hight density polyethylene), PVC (Polyvinyl chloride), LDPE (Low density polyethlene), PP (Polyethylene), PS (Polyethlene) and Other (O) arise due to a lack of understanding of the different types of plastic and how to identify them. This knowledge is especially important in the context of sustainable plastic waste management. In an effort to address this challenge, object detection and classification technologies are gaining increasing attention in the recognition and management of various types of plastics. This paper proposes a detection and classification approach to identify 7 (seven) types of plastic that are commonly used using object detection techniques, with a focus on using YOLO (You Only Look Once) v8, a new method in computer vision. The proposed method is implemented by utilizing the YOLOv8 architecture, YOLOv8 is the latest version which has several improvements and advantages for recognizing and distinguishing objects including types of plastic in image form. The dataset covering a wide variety of plastics is used as the basis for training detection and classification models. The results of the experiments demonstrate the effectiveness and accuracy of the proposed approach in recognizing and classifying plastics based on their visual features.