YOLOv8 for Real-Time Microplastic Detection in Aquatic Environments

Abstract: Plastic waste is destroying marine ecosystems throughout the world and posing a threat to the life of animals and the health of the waters. This does research on a new deeplearning model named Yolov8 to identify and classify underwater plastic trash. The system is highly functional in all sorts of water, and can detect different plastic waste in real time. The technology can be useful in the context of conserving the ocean by providing more accurate estimates of the amount of trash that exists, its origin, and the optimal clean-up strategies. Testing indicates that accuracy of the new method is far much superior to the previously used methods particularly when visibility of the water is low. The article contributes to the emerging research on AI in environmental monitoring, by offering a scaled instrument and measuring plastic pollution to direct conservation and policy-related actions.

Keywords: Locating plastic, gathering garbage, categorizing them, YOLOv8, computer vision, and deep learning.