Effective reduction of oceanic plastic pollution requires scalable and objective monitoring methods that go beyond traditional human-based surveys. This review synthesizes recent advances in remote sensing and AI-driven image analysis for detecting macro-plastic litter. Peer-reviewed studies published up to 2024 were systematically selected from the Scopus database, focusing on applications of remote sensing platforms including webcams, drones, balloons, aircraft, and satellites for monitoring plastic litter in coastal, riverine, and other aquatic environments. Quantification methods ranged from manual annotation to deep learning-based models. Although machine learning has been increasingly adopted since around 2020, manual screening and rule-based approaches remain prevalent, reflecting the complexity of litter types and shapes. The review revealed considerable variability in quantification metrics-such as litter-covered area, volume, weight, and item count per unit area-which complicates cross-study comparisons and data harmonization. While remote sensing enhances spatial coverage, consistency, and repeatability, it faces persistent challenges, including environmental interference, limited resolution, and inconsistent protocols. Our findings highlight the urgent need for methodological standardization and harmonization of quantification units across platforms and geographic regions. Among available metrics, litter-covered area and item count per unit area are most suitable for cross-platform comparison. Continued development and integration of such technologies hold strong potential to facilitate science-based policymaking and long-term monitoring of plastic transport from land to ocean.
Kako et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: