Plastic image detection from UAV images using YOLOV8

Plastic pollution is a critical environmental issue that demands innovative monitoring solutions. This study utilizes the YOLOv8 deep learning algorithm in combination with Unmanned Aerial Vehicles (UAVs) equipped with high-resolution cameras to automate plastic detection in aerial imagery. A custom drone dataset, with 85% of the images containing plastic, was developed to evaluate detection performance. The results demonstrate that YOLOv8 can effectively identify plastic in complex environments, offering a scalable and efficient alternative to traditional ground surveys. However, challenges such as varied plastic types, sizes, and visual complexity remain. This UAV-based approach shows promise for large-scale applications, including tracking cleanup initiatives, mapping plastic distribution, and assessing environmental impacts. Future research could improve this methodology by incorporating advanced image processing techniques to address plastic variability and enhance detection accuracy. Overall, the findings highlight UAVs as a valuable tool for comprehensive plastic pollution monitoring and contribute to advancing environmental conservation efforts.