International Journal of Engineering and Manufacturing (IJEM)
IJEM Vol. 16, No. 4, 8 Aug. 2026
Cover page and Table of Contents: PDF (size: 849KB)
Hybrid ViT-UNet Framework for Accurate River Segmentation and Buffer Zone Mapping in High-Resolution Satellite Imagery
PDF (849KB), PP.71-80
Views: 0 Downloads: 0
Author(s)
T. Satyanarayana Murthy
1
K. Gangadhara Rao
1
Swathi Sowmya Bavirathi
1
Index Terms
Segmentation, Remote Sensing, Vision Transformers, Flood Prediction, Semantic Segmentation
Abstract
Precise mapping of water bodies is crucial for flood monitoring, disaster risk and response reduction, as well as sustainable water resource management. In this paper, we introduce a deep learning model for effective segmentation of rivers, lakes, and reservoirs from high-resolution Gaofen-2 satellite images. Leveraging the Five-Billion-Pixels dataset-more than 5 billion annotated pixels for 24 land cover classes—our approach solves the problem of segmenting water bodies on various terrains and environmental conditions. The proposed U-Net and ViT-UNet models, with the former employing Vision Transformers to enhance global context perception. For enhancing generalization, the dataset is augmented using Albumentations and flipping, rotation, and scaling transformations. Hybrid loss functions of Dice Loss, Binary Cross-Entropy, and Focal Loss are employed to handle class imbalance, especially for slender river segments. The ViT-UNet model attained 98.8% pixel accuracy, which mirrors its ability to preserve fine detail and large-scale spatial pattern. Mixed-precision training and the AdamW optimizer has enhanced the computational efficiency. Further, demonstrates the potential of transformer-based segmentation models for remote sensing achieved accuracy of 98% for environmental risk management and decision support in disaster-prone areas.
Cite This Paper
T. Satyanarayana Murthy, K. Gangadhara Rao, Swathi Sowmya Bavirathi, "Hybrid ViT-UNet Framework for Accurate River Segmentation and Buffer Zone Mapping in High-Resolution Satellite Imagery", International Journal of Engineering and Manufacturing (IJEM), Vol.16, No.4, pp.71-80, 2026. DOI:10.5815/ijem.2026.04.05
Reference
[1]Bui, T.A. and Lee, P.J., 2023. SWRNet: A deep learning approach for Small surface Water area Recognition onboard satellite. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 16, pp.10369-10380.doi: 10.1109/JSTARS.2023.3317268.
[2]Li, Y., Zhu, Z., Li, Y., Zhang, J., Li, X., Shang, S. and Zhu, D., 2023. Ctmu-net: an improved u-net for semantic segmentation of remote-sensing images based on the combined attention mechanism. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 16, pp.10148-10161.doi: 10.1109/JSTARS.2023.3316725.
[3]Lin, X., Cheng, Y., Chen, G., Chen, W., Chen, R., Gao, D., Zhang, Y. and Wu, Y., 2023. Semantic segmentation of China’s coastal wetlands based on Sentinel-2 and Segformer. Remote Sensing, 15(15), p.3714.doi: 10.3390/rs15153714.
[4]Sun, D., Gao, G., Huang, L., Liu, Y. and Liu, D., 2024. Extraction of water bodies from high-resolution remote sensing imagery based on a deep semantic segmentation network. Scientific Reports, 14(1), p.14604.doi: 10.1038/s41598-024-95427-9.
[5]Chen, X., Li, D., Liu, M. and Jia, J., 2023. CNN and transformer fusion for remote sensing image semantic segmentation. Remote Sensing, 15(18), p.4455.doi: 10.3390/rs15184455.
[6]Ghaznavi, Ali, et al. "Comparative performance analysis of simple U-Net, residual attention U-Net, and VGG16-U-Net for inventory inland water bodies." Applied Computing and Geosciences 21 (2024): 100150.doi: 10.1016/j.acags.2024.100150.
[7]Yin, Y., Guo, Y., Deng, L. and Chai, B., 2023. Improved PSPNet-based water shoreline detection in complex inland river scenarios. Complex & Intelligent Systems, 9(1), pp.233-245.doi: 10.1007/s40747-022-00869-9.
[8]Mayer, Timothy, et al. "Deep learning approach for Sentinel-1 surface water mapping leveraging Google Earth Engine." ISPRS Open Journal of Photogrammetry and Remote Sensing 2 (2021): 100005.doi: 10.1016/j.ophoto.2021.100005.
[9]Pena, Francisco J., et al. "Deepaqua: Self-supervised semantic segmentation of wetland surface water extent with sar images using knowledge distillation." arXiv preprint arXiv:2305.01698 (2023).doi: 10.48550/arXiv.2305.01698.
[10]Billson, J., Islam, M.S., Sun, X. and Cheng, I., 2023. Water body extraction from sentinel-2 imagery with deep convolutional networks and pixelwise category transplantation. Remote Sensing, 15(5), p.1253.doi: 10.3390/rs15051253.
[11]Silpalatha, G., and T. S. Jayadeva. "Comprehensive study on the water body and environmental issues–A review." AIP Conference Proceedings. Vol. 3028. No. 1. AIP Publishing, 2024.doi: 10.1063/5.0183647.
[12]Chen, Y., Fan, R., Yang, X., Wang, J. and Latif, A., 2018. Extraction of urban water bodies from high-resolution remote-sensing imagery using deep learning. Water, 10(5), p.585.doi: 10.3390/w10050585.
[13]Z. Feng, F. Zhang, X. Ma, C. Y. Jim, D. Wu, D. Liu, S. A. Oke, W. Wang, L. Wei, S. Nie, Z. Li, and H. Xu, “Refining water body extraction by remote sensing with deep learning models: Exploring different band combinations,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 18, pp. 18005–18018, 2025. doi: 10.1109/JSTARS.2025.3432761.
[14]Xiaofei Han, Nazih Y. Rebouh, Yasmeen Ahmed, Muhammad Nasar Ahmad, Zainab Tahir, Yahia Said, and Ishfaq Gujree “Enhancing water bodies detection in the highland and coastal zones through multi-sensor spectral data fusion and deep learning,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 18, pp. 15629–15643, 2025. doi: 10.1109/JSTARS.2025.3424128.
[15]J. Wang, D. Jia, J. Xue, Z. Wu, and W. Song, “Automatic water body extraction from SAR images based on MADF-Net,” Remote Sensing, vol. 16, no. 18, p. 3419, 2024. doi: 10.3390/rs16183419.
[16]L. Lou, C. Chen, M. Li, and K. Liu, “Comparative analysis of water body extraction accuracy based on thresholding method,” ISPRS Archives, vol. XLVIII-4, pp. 331–336, 2024. doi: 10.5194/isprs-archives-XLVIII-4-331-2024.
[17]Q. Teng, L. Luo, S. Li, L. Xing, K. Shao, S. Wang, and D. Wang, “A framework for flood inundation extraction based on microwave and optical remote sensing images,” Frontiers in Earth Science, vol. 12, p. 1511834, 2024. doi: 10.3389/feart.2024.1511834.