• Ms. Jami Kavitha
• A. John
• V. Chandu
• J. Rushitha
• P. Akash

Flooding is a significant natural disaster that causes widespread damage, making rapid and accurate identification of affected areas essential for efficient disaster response and management. Satellite imagery provides a valuable resource for detecting flood-affected regions, but manually identifying flooded areas from large-scale satellite images is both time-consuming and prone to errors. This paper proposes an approach for automating flood area segmentation using SegNet, a deep learning architecture designed specifically for semantic segmentation tasks. By employing SegNet’s encoder-decoder structure, which preserves spatial resolution and boundary details, we achieve more accurate flood area detection compared to conventional Fully Convolutional Networks (FCNs). Our model is trained and validated on a high-resolution satellite imagery dataset, showcasing significant improvements in segmentation accuracy. Results demonstrate the effectiveness of SegNet in identifying flood-affected areas, providing a robust tool for disaster management. This work has important implications for real-time monitoring, response strategies, and decision-making in flood-prone regions.

Flood segmentation, SegNet, deep learning, semantic segmentation, disaster management, satellite imagery, image analysis, encoder-decoder model, flood detection, spatial resolution.

"FLOOD AREA SEGMENTATION USING SEGNET", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 4, page no.f268-f275, April-2025, Available :http://www.jetir.org/papers/JETIR2504536.pdf

"FLOOD AREA SEGMENTATION USING SEGNET", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 4, page no. ppf268-f275, April-2025, Available at : http://www.jetir.org/papers/JETIR2504536.pdf