We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Modern Floating Marine Macrolitter Monitoring Approaches and Methods and Integration of Machine Learning Techniques
AI summary Read the abstract
This review examined modern methods and approaches for detecting floating marine macrolitter, including visual observations, trawling, radar remote sensing, and deep learning techniques. The authors analyzed recent advances in artificial neural networks and machine learning that have significantly improved the accuracy of identifying and classifying marine litter at the sea surface, with key challenges including object diversity and partial submersion.
Marine litter pollution is currently recognized as global problem at the level of all international organizations and conventions related to the marine protection. This review examines modern methods and approaches for detecting floating marine macrolitter. The task of detecting marine litter on the water surface is complicated by a large variety of objects, various degrees of their degradation, predominantly small size, partial immersion in the subsurface layer, colorlessness, disguising within the water, difficult observation conditions. The main approaches today include visual observations (from ships, aircraft), trawling, and remote sensing, especially using radar systems. In the last decade, deep learning methods have made significant progress, which has allowed error recognition and identification to be brought to a new level due to various modifications of artificial neural networks. In this review, we analyze the main research on the presented topic and significant achievements and prospects for the application of artificial intelligence to improve methods for detecting and classifying marine litter larger than 2.5 cm.
More Papers Like This
State of the art applications of deep learning within tracking and detecting marine debris: A survey
AI summary Read the abstract
This survey analyzed 28 recent deep learning studies applied to marine debris detection and tracking, finding that YOLO-based models significantly outperform other object detection methods. Testing on a small curated dataset revealed low accuracy and high false positive rates, underscoring the urgent need for a comprehensive annotated database of underwater plastic debris.
Exploratory data analysis of visual sea surface imagery using machine learning
AI summary Read the abstract
An anomaly detection algorithm based on contrastive learning and supervised machine learning successfully identified floating marine litter, birds, and unusual glare in optical sea surface imagery. Scalable, automated surface monitoring tools are critical for tracking plastic debris distribution and density across the open ocean where manual surveys are impractical.
Deep-Feature-Based Approach to Marine Debris Classification
AI summary Read the abstract
This study applied deep learning to classify marine debris from images, demonstrating that feature-based neural network approaches can effectively distinguish plastic types and other debris categories to support automated ocean monitoring.
Up-to-Date Scoping Review of Object Detection Methods for Macro Marine Debris
AI summary Read the abstract
This review examines over two decades of research into object detection methods for identifying macro marine debris, with a focus on deep learning advances made in the past five years. Following PRISMA-ScR guidelines, the authors assessed current techniques and identified key progress and remaining gaps in automated marine litter identification systems.
Remote sensing and image analysis of macro-plastic litter: A review
AI summary Read the abstract
This systematic review synthesized recent advances in remote sensing and AI-driven image analysis for detecting macro-plastic litter in coastal, riverine, and aquatic environments using platforms from webcams to satellites. The review found considerable variability in quantification metrics and highlighted the urgent need for methodological standardization to enable cross-study comparisons and evidence-based policymaking.
Research digests by email
When a large batch of papers lands in the Atlas, we read through it and send a short write-up of what stood out.