Efficient Identification of Waste in Water Bodies with YOLO V7-Object Detection

Marine pollution is made up of waste and debris, the majority of which comes from land and is thrown or blown into the water. The ecology, the health of all living things, and international financial organizations are all negatively impacted by this pollution. The threat of marine pollution to civilization is one of the biggest. All manufactured items, the majority of which are made of plastic and end up in the water, are considered marine trash. This material builds up as a result of littering, storm gusts, and poor waste management, and in 80% of the cases, it originates from sources on land. In this study, we suggest an improved object identification model based on deep learning to find waste in water bodies and protect marine life. In order to locate waste in water bodies and save marine life, we propose an improved deep learning-based object detection model to address this problem. YOLOv7 performs faster and more accurately than any other object detectors between 5 and 160 frames per second. Large datasets of images of marine trash are used to train this model and is able to accurately identify a wide variety of objects, including plastic bags, bottles, straws, and other debris.