Development of automated marine floating plastic detection system using Sentinel-2 imagery and machine learning models

Abstract

The increasing level of marine plastic pollution poses severe threats to the marine ecosystem and biodiversity. Open remote sensing data and advanced machine learning (ML) algorithms could be a cost-effective solution for identifying large plastic patches across the scale. The potential application of such resources in detecting and discriminating marine floating plastics (MFP) are not fully explored. Therefore, the present study attempted to explore the full functionality of open Sentinel satellite data and ML models for detecting and classifying the MFP in Mytilene (Greece), Limassol (Cyprus), Skala Loutron, Greece, Calabria (Italy), and Beirut (Lebanon). Two ML models, i.e. Support Vector Machine (SVM) and Random Forest (RF), were utilized to perform the classification analysis. In-situ plastic location data was collected from the control experiments conducted in Mytilene, Greece (in 2018 and 2019), Skala Loutron, Greece (2021), and Limassol, Cyprus (2018), and the same was considered for training the models. The accuracy and performances of the trained models were further tested on unseen new data collected from Calabria, Italy and Beirut, Lebanon. Both remote sensing bands and spectral indices were used for developing the ML models. A spectral signature profile for marine plastic was created for discriminating the floating plastic from other marine debris. A newly developed index, kernel Normalized Difference Vegetation Index (kNDVI), was incorporated into the modelling to examine its contribution to model performances. Both SVM and RF were performed well in five models and test case combinations. Among the two ML models, the highest performance was measured for the RF. The inclusion of kNDVI was found effective and increased the model performances, reflected by high balanced accuracy measured for model 2 (~89% to ~100% for SVM and ~92% to ~98% for RF). An automated floating plastic detection system was developed and tested in Calabria and Beirut using the best-performed model. The trained model had detected the floating plastic for both sites with ~80%-90%% accuracy. Among the six predictors, the Floating Debris Index (FDI) was the most important variable for detecting marine floating plastic. These findings collectively suggest that high-resolution remote sensing imagery and the automated ML models can be an effective alternative for the cost-effective detection of MFP. Future research will be directed toward collecting quality training data to develop robust automated models and prepare a spectral library for different plastic objects for discriminating plastic from other marine floating debris and advancing the marine plastic pollution research by taking full advantage of open-source data and technologies.

Keywords: Machine learning; Marine litter; Open data; Plastic pollution; Remote sensing; Sentinel.