We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Plastic Litter Project 2018: a feasibility study on detecting plastics in the aquatic environment using drones and satellite images
Summary
Researchers tested whether drone and satellite imagery could detect floating plastic litter in the ocean, using large artificial plastic targets made of PET bottles, plastic bags, and nylon fishing nets. While drones could detect the targets, satellite resolution was insufficient at the tested target size, highlighting current limitations of remote sensing for marine plastic litter monitoring.
Plastic litter has been shown to pose a significant problem in the marine environment and the food chain at all trophic levels. Within the scope of Plastic Project 2018, we investigated the prospective use of imaging technology on airborne and spaceborne platforms in detecting floating plastic litter. Three artificial targets of plastic litter were prepared using PET-1water bottles, LDPE plastic bags and nylon fishing ghost nets. Each target measured 10 m x 10 m fixed by a PVC frame to prevent spillage as well as to make an aggregated target. We evaluated the derived spectral reflectances of these plastic litter targets gathered from the airborne (drone) and spaceborne (Sentinel-2) images. Improved geo-referencing of the high geo-spatial resolution (resampled 10 m) Sentinel-2 images was achieved by utilizing excellent geospatial resolution (<0.03 m) drone images. Further analysis involved determining accurate pixel coverage of each target with future application in quantification efforts. Of the three plastic litter targets, it was noted that the plastic bottles had the highest spectral reflectances that were measured by the Sentinel-2 mission image. The experiment showed that Sentinel-2 satellites could be used to detect marine plastics when they cover a large area. The Sentinel-2 NIR band can significantly contribute to marine litter detection and the water content partly influence the behaviour of spectral values.