River Area Segmentation Using Sentinel-1 SAR Imagery with Deep-Learning Approach

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Deep-learning models enhance river segmentation using Sentinel-1 SAR imagery, indicating improved mapping capabilities.

CNN U-Net achieved an accuracy of 0.94 for river segmentation, outperforming DeepLabv3+ which attained 0.92.
River segmentation was improved through deep-learning-based models, utilizing Sentinel-1 SAR imagery for precise data extraction.
The study trained the models on a total of 1080 SAR images, leveraging the unique capabilities of C-Band radar technology.
Enhanced river mapping enables better management for irrigation and flood-monitoring applications in tropical regions.

Dewi et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark