While preventing the flow of plastic waste into the ocean is key to safeguarding the health of our marine environments, it’s also essential that we clean up the mess that’s already there. Of course, detecting ocean pollution is not an easy undertaking — but a new useful tool is here to help.
Developed by researchers at the University of Barcelona, the new method involves an AI algorithm that can detect and quantify litter through aerial imagery, and could soon tap into the potential of drones to autonomously scan the seas and assess the damage.
With tons of plastic waste leaking into our oceans on a daily basis and so much of it breaking down into smaller fragments, taking stock of our plastic pollution problem is rather difficult. That’s exactly what has driven the University of Barcelona team to take aim at the problem by improving current techniques that track and trace smaller pieces of plastic floating on the surface.
As part of the process, the team turned to deep learning techniques to analyze more than 3,800 aerial images of the Mediterranean off the coast of Catalonia. By feeding these photographs to the algorithm and using neural networks to improve accuracy over time, the team wound up with an AI tool that can reliably detect and measure the amount of plastic floating on the surface.
“The great amount of images of the marine surface obtained by drones and planes in monitoring campaigns on marine litter — also in experimental studies with known floating objects — enabled us to develop and test a new algorithm that reaches an 80 percent of precision in the remote sensing of floating marine macro-litter,” says team member Odei Garcia-Garin.
The tool can identify individual plastic items as well as categorize them accordingly. And since it’s an open-access web app, the software can be used by professionals in the field. The team is also currently working on a version that could work with drones, to fully automate the process.
