Real-Time Oil Spill Detection Using U-Net Segmentation on SAR Imagery from Sentinel and PALSAR with Severity Alerts via AWS SNS

DOI

10.2991/978-94-6463-978-0_13How to use a DOI?

Keywords

Oil spill detection; SAR imaging; Sentinel-1; PALSAR; U-Net; Severity assessment; AWS SNS; Real-time monitoring

Abstract

Oil spills have long posed serious threats to marine ecosystems and the coastal economies that rely on them. While early identification is essential for limiting damage, detecting spills under constantly changing ocean conditions remains difficult. Synthetic Aperture Radar (SAR) imagery is especially useful in this regard because it can capture sea-surface characteristics regardless of weather or daylight. However, conventional approaches such as thresholding or texture-based methods often mistake natural ocean phenomena for actual spills, requiring human verification to confirm the results.

In this study, we develop a lightweight U-Net segmentation model trained on SAR data obtained from the Sentinel-1 (C-band) and PALSAR (L-band) satellites, aiming to improve detection reliability across varying sea states. The model estimates the affected spill area through pixel-level spatial analysis and classifies each event into three severity categories. To support rapid response, the framework integrates an event-driven alert system using\ AWS Simple Notification Service (SNS), which automatically notifies responsible authorities when new detections occur.

The proposed system achieved 86.2% accuracy, with IoU = 79.5% and Dice = 82.8%, demonstrating consistent segmentation performance. A user-friendly Streamlit dashboard enables real-time visualization of spill regions and severity levels. Together, these components provide a scalable and practical solution for automated oil spill detection, monitoring, and early decision support in marine environments.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - V. Padmapriya
AU  - S. K. Theekshanaa
AU  - Yuvaraj Natarajan
PY  - 2025
DA  - 2025/12/31
TI  - Real-Time Oil Spill Detection Using U-Net Segmentation on SAR Imagery from Sentinel and PALSAR with Severity Alerts via AWS SNS
BT  - Proceedings of the 1st Engineering Data Analytics and Management Conference (EAMCON 2025)
PB  - Atlantis Press
SP  - 129
EP  - 144
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-978-0_13
DO  - 10.2991/978-94-6463-978-0_13
ID  - Padmapriya2025
ER  -