Japan research team develops AI system to detect plastic litter on seabed
TOKYO (Kyodo) -- A Japanese team of researchers has developed an artificial intelligence system to better detect marine plastic waste on the seabed, i
Japanese scientists have unveiled an artificial intelligence (AI) system designed to more accurately detect plastic waste on the ocean floor, according to a recent study published in an international journal. The system, named DeepLitterAI, was developed by researchers at the Japan Agency for Marine-Earth Science and Technology.
The AI system boasts the ability to process data at nearly double the speed of human analysis, making it a promising tool for real-time monitoring of marine plastic pollution. To train the system, researchers compiled a dataset comprising roughly 12,000 images of Japan's seabed collected by the agency since 1983. This dataset included images of both litter and non-litter objects that can be easily misidentified, such as rocks and marine life.
To minimize false detections, the team subjected the dataset to a variety of processing techniques, including blurring and inversion, to simulate different conditions the AI might encounter in real-world scenarios. Ryota Nakajima, a biological oceanographer involved in the project, explained that the system could quickly pinpoint areas with significant litter accumulation, enabling authorities to implement targeted countermeasures.
During tests using actual footage, DeepLitterAI successfully identified the type and quantity of litter even when it comprised only 5 to 10 percent of the image width. The system accurately detected 80 percent of major litter items, such as plastic bottles and polythene bags, with an average margin of error of around 10 percent compared to visual inspections by experts. Importantly, the AI could complete an analysis that would typically take a human expert about a month in just a few days.
The researchers note that this innovation addresses a limitation of existing AI systems, which often struggle to recognize smaller plastic items due to the wide-angle lenses used in deep-sea footage. By improving accuracy by 1.6 times compared to previous methods, DeepLitterAI represents a significant step forward in the fight against marine plastic pollution.
Written by urgent.news from The Mainichi's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
