Iris Coleman
June 8, 2025 02:33
New AI tools have been developed to identify and find abandoned fishing nets known as ghost nets using advanced underwater video technology.
In a significant progress for marine preservation, a new AI tool has been developed to identify and find abandoned fishing nets called ghost nets. This technology aims to alleviate the harmful effects of this net on marine wildlife. According to NVIDIA’s blogs, preservators have introduced a platform that can analyze vast amounts of underwater video data to accurately point these ghost nets.
Environmental influence of the ghost net
About 2%of the world’s fishing equipment, including about 80,000 square kilometers of fishing nets every year, is lost in the sea. This net leads to a serious threat to marine wildlife such as seals, turtles and dolphins, which often leads to entanglement and subsequent drowning. In addition, when this net is decomposed, it will greatly contribute to plastic contamination in the marine environment.
Technology innovation of detection
Often, the detection of this ghost net, which is as thin as a finger, presents a significant challenge. To solve this, WWF Germany developed Ghostnetzero.ai in cooperation with AI for AI for Accenture and Microsoft. This online platform counts high -resolution underwater data from various sources, including research institutes and overseas wind power companies, to help the ghost net identification.
The AI tool uses the Convolutional Neural Network (CNN) with DEEPLABV3 and Resnet50 50 backs to scan the cattle and data on the signs of the ghost net to achieve impressive 94% accuracy. This precision contains geographic location metadata in each cow or image, so you can find exactly the ghost net location.
Cooperation efforts to remove net removal
When the ghost nets are confirmed, local divers and fishermen participate in to confirm their existence and to collect more information as the size of the net. Such insights are important for organizing logistics necessary for net removal. The WWF German project manager, Gabriele Dederer, emphasizes the importance of local collaboration and states that regional protocols are essential due to logistics and financial complexity related to searching this net.
This initiative is currently planning to expand further in countries such as France, Estonia and Sweden. Integrating AI into this project is promoted by Microsoft’s cloud service and NVIDIA’s GPU technology, improving the accuracy and efficiency of the model.
Future prospects and technology development
Christian Bucher, a representative of Microsoft, who participated in this project, emphasized the role of AI in pattern recognition. This project continues to develop as local team verification returns to the AI model and improves performance.
This innovative approach to resolving marine pollution show promising progress in global efforts to protect the marine ecosystem. By utilizing state -of -the -art technology, the initiative not only identifies threats, but also opens up the way for effective removal and preventive strategies.
Image Source: Shutter Stock