Deep Learning-Based Flood Area Segmentation with CBAM-Enhanced DeepLabV3+

Natural calamities like floods have a significant impact on the environment, society, and the economy. It can result in huge financial losses, community displacement, infrastructure ruin, and fatalities. Timely and precise flood detection is vital for effective disaster management and limiting the adverse impacts of floods. The accurate identification of flood-affected water regions remains a challenging task due to variations in water appearance, vegetation, urban structures, and complex terrain. This study highlights the significance of defining water zones in flood-affected regions effectively. This study investigates the use of deep learning models like UNet and DeepLabV3 + for automatic segmentation of flood areas. Convolutional Block Attention Module (CBAM) has been employed in this study to focus on the most relevant spatial and channel-wise features. The use of CBAM enhances the model’s ability to accurately identify flood-affected areas. To train, test and validate the proposed model, a dataset containing 290 images representing flood-affected areas is obtained from Kaggle. The study evaluates UNet and DeepLabV3 + deep learning models with ResNet50 and ResNet101 backbones architectures for flood-affected area segmentation. The proposed DeepLabV3 + model with CBAM and ResNet101 as backbone architecture achieved the best performance, with an average IoU of 0.8889, DICE score of 0.9391, and precision of 0.9426. These findings highlight the potential of deep learning approaches for automated flood-area segmentation, which can enhance flood monitoring, disaster assessment, and rescue operations.

Authors

Institutions

Publication Details

Journal
Journal of Transformative Technologies and Sustainable Development
Published
2026-11-16
DOI
https://doi.org/10.1007/s41314-026-00100-0
Primary Topic
Flood Risk Assessment and Management
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Deep Learning-Based Flood Area Segmentation with CBAM-Enhanced DeepLabV3+

Swapandeep Kaur, Priyanka Malhotra
4 links
Journal of Transformative Technologies and Sustainable Development
Flood Risk Assessment and Management
article

Deep Learning-Based Flood Area Segmentation with CBAM-Enhanced DeepLabV3+

Swapandeep Kaur, Priyanka Malhotra
article en

Abstract

Natural calamities like floods have a significant impact on the environment, society, and the economy. It can result in huge financial losses, community displacement, infrastructure ruin, and fatalities. Timely and precise flood detection is vital for effective disaster management and limiting the adverse impacts of floods. The accurate identification of flood-affected water regions remains a challenging task due to variations in water appearance, vegetation, urban structures, and complex terrain. This study highlights the significance of defining water zones in flood-affected regions effectively. This study investigates the use of deep learning models like UNet and DeepLabV3 + for automatic segmentation of flood areas. Convolutional Block Attention Module (CBAM) has been employed in this study to focus on the most relevant spatial and channel-wise features. The use of CBAM enhances the model’s ability to accurately identify flood-affected areas. To train, test and validate the proposed model, a dataset containing 290 images representing flood-affected areas is obtained from Kaggle. The study evaluates UNet and DeepLabV3 + deep learning models with ResNet50 and ResNet101 backbones architectures for flood-affected area segmentation. The proposed DeepLabV3 + model with CBAM and ResNet101 as backbone architecture achieved the best performance, with an average IoU of 0.8889, DICE score of 0.9391, and precision of 0.9426. These findings highlight the potential of deep learning approaches for automated flood-area segmentation, which can enhance flood monitoring, disaster assessment, and rescue operations.

Journal of Transformative Technologies and Sustainable Development
Chitkara University (IN)
Clean water and sanitation
Flood Risk Assessment and Management
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon