What question did this study set out to answer?

This research aims to compare the performance of different deep learning architectures for accurately segmenting water bodies from satellite images.

May 31, 2026Open Access

GeoAI Framework for Evaluation of Deep Learning Models in Water Body Segmentation

Key Points

This research aims to compare the performance of different deep learning architectures for accurately segmenting water bodies from satellite images.
Evaluated four deep learning architectures: U-Net, U-Net++, DeepLabV3+, and Feature Pyramid Network using ResNet-34 encoder.
Applied consistent preprocessing, data augmentation, and training protocols across all models on the Earth Surface Water Dataset.
Determined segmentation performance using Intersection over Union (IoU) and Dice coefficient metrics.
DeepLabV3+ achieved the highest accuracy with an IoU of 0.955 and Dice of 0.973.
U-Net++ came next with an IoU of 0.945 and Dice of 0.969, followed by U-Net with an IoU of 0.9352 and Dice of 0.9598.
FPN's performance was lower, with an IoU of 0.9218 and Dice of 0.9514.

Abstract

Abstract. Accurate delineation of surface water bodies from satellite imagery is fundamental for hydrological monitoring, disaster management, and environmental assessment. Deep learning–based semantic segmentation has shown great promise for this task, yet the relative performance of widely used encoder–decoder architectures remain insufficiently explored under consistent experimental settings. This study presents a GeoAI-based evaluation of four state-of-the-art segmentation architectures: U-Net, U-Net++, DeepLabV3+, and Feature Pyramid Network (FPN), all employing a ResNet-34 encoder pretrained on ImageNet. Using the Earth Surface Water Dataset, we applied an identical tiling strategy, preprocessing pipeline, data augmentation, optimizer configuration, and training schedule to ensure a fair comparison. Our results show that DeepLabV3+ achieved the highest accuracy with an IoU of 0.955 and Dice of 0.973, followed by U-Net++ (IoU 0.945, Dice 0.969) and U-Net (IoU 0.9352, Dice 0.9598). FPN performed comparatively lower with an IoU of 0.9218 and Dice of 0.9514. These findings demonstrate that architectural choices substantially influence segmentation quality even with identical encoders and training protocols. The study provides benchmark results and practical insights for selecting deep learning architectures for operational water body mapping in remote sensing applications.

Ask AI

Helpful

Bookmark

View Full Paper