MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641093614 A) filed by Easwari Engineering College; and Srm Institute Of Science And Technology, Ramapuram Campus on August 03, 2026, for An Automated Deep Learning System For Marine Oil Spill Detection, Segmentation And Spread Prediction From Sentinel-1 Sar Imagery.
Inventors include Kishore V; Krishnakumar R; and Dr. Meena Azhagiri.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: Oil spills pose significant threats to marine ecosystems, requiring rapid and accurate detection for effective mitigation. The present invention introduces SpillXNet, an AI-based framework for automated oil spill detection, segmentation, and spread analysis using Synthetic Aperture Radar imagery. The system uses dual-polarized SAR (VV and VH) images through normalization, resizing to 512 × 512 pixels, and conversion into optimized arrays for efficient training. SpillXNet, a lightweight CNN, classifies SAR images into Oil Spill, No Spill, and Lookalike categories, minimizing false detections caused by lookalike phenomena. Upon oil spill detection, a U-Net segmentation model employing a composite Binary Cross-Entropy and Dice Loss accurately delineates spill boundaries. The segmented output is further processed by a spread prediction module that estimates spill movement using parameters, including ocean current speed, ocean current direction, wind speed, and wind direction. The framework provides an end-to-end intelligent solution for real-time monitoring, environmental surveillance, and emergency response. Total Number of Words in Abstract:149
Disclaimer: Curated by HT Syndication.
