We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Automated Marine Debris Detection and Segmentation Using Yolov11 and Deeplabv3 for Marine Imagery
Summary
This study combined YOLOv11 for real-time object detection and DeepLabV3+ for semantic segmentation to autonomously detect and segment plastic marine waste from satellite imagery. The models were trained on the Marine Debris v49 dataset from Roboflow and MARIDA archives, using stratified splits and augmentation techniques to address challenges in monitoring dynamic marine plastic pollution environments.
Marine plastic debris represents a vital danger to ocean ecosystems, maritime navigation safety, and human lives.Traditional monitoring and cleanup techniques are time-consuming, labor-intensive, and often fail to capture the dynamic character of marine environments.This research introduces an autonomous deep learning system that combines YOLOv11 for real-time object detection and DeepLabV3+ for semantic segmentation of plastic marine waste from satellite images.The models learned on the Marine Debris v49 dataset, which was filtered from Roboflow and MARIDA archive, utilizing stratified splits to maintain class balance.The data preprocessing involved resizing, normalization, and augmentation techniques such as random rotation,