Marine plastic litter has become a global environmental problem with impacts on marine biodiversity and human health. VITO Remote Sensing is using its knowledge on in-water optics and water quality retrieval from remote sensing to provide solutions for the marine plastic litter problem.
Both spectral approaches as artificial intelligence are being used to detect and characterize the marine plastic macro litter from drones, fixed cameras and satellite observations. These observations can provide standardized means to follow up regulations or they can be linked to clean-up activities.
VITO initiated the HYPER project in 2019 and built up a spectral database of different types of plastics, including wet, dry and submerged plastics, both virgin and weathered samples. The project has resulted in a database which is publicly available and several peer reviewed publications have been submitted. This knowledge is used to develop machine learning algorithms to detect individual pieces of litter in waterways and litter accumulations on land.
In the video below, our expert in water & coastal monitoring Els Knaeps stresses the importance of these measurements and explains more about the difficulties when detecting plastics in real marine conditions.
The left image is showing the input data. The image on the right shows the result of the bottles found using Artificial Intelligence.
Related projects:
- HYPER - Hyperspectral Remote Sensing of Marine Plastics (H2020 ATTRACT)
- PLUXIN - Plastic Flux for Innovation and Business Opportunities in Flanders (Blue Cluster)
- AIDMAP - Artificial Intelligence and drones supporting the detection and mapping of floating aquatic plastic litter (ESA)
- MARLISE - Marine Litter Mission Conceptual Design Study (ESA PRODEX)
Related papers and datasets:
- Spectral reflectance of marine macroplastics in the VNIR and SWIR measured in a controlled environment
- Hyperspectral-reflectance dataset of dry, wet and submerged marine litter
- Hyperspectral Remote Sensing of Marine Plastics (HYPER) - PDF download
- Dataset of different types of plastics - click here to download
- Targeting Plastics: Machine Learning applied to litter detection in aerial multispectral images
Related blog article: