Proceedings of the First International Conference on Artificial Intelligence, Smart Technologies and Communications (AISTC 2025)

<Previous Article In Volume
Next Article In Volume>

New Architecture of Deep Learning using DeepLabV3+ ResNet50 and ResNet18 to Extract Water Bodies

Authors
Abdelali Benali1, *, Hayet Kharbouch2, Yesma Bendaha1
1Automatic Department, University of Sciences and Technology of Oran Mohamed BOUDIAF, 1505, El M’naouer Oran, Algeria
2University of Science and Technology MB, Laboratory of Applicated Power Electronic (LEPA), Oran, Algeria
*Corresponding author. Email: benabdel0305@gmail.com
Corresponding Author
Abdelali Benali
Available Online 5 August 2025.
DOI
10.2991/978-94-6463-805-9_15How to use a DOI?
Keywords
Deep-learning; Water-bodies; Deeplabv3+; ResNet; Classification
Abstract

The inventory of water surfaces is vital for the survival of the human race and the preservation of the ecosystem. Several approaches and methods have been developed to extract water bodies, the principal tool currently used for this task is deep learning.

The major problem is the need for a large amount of information used for training, and the big computational time.

The most used satellite image dataset is the Sentinel-2 Dataset that we down - loaded from the Kaggle. This image dataset contains several images of different resolutions, dimensions, textures, and complexity.

The originality of our work was to choose the best architecture of the model, we used two combinations between the DeepLabv3 +, ResNet50, and ResNet18 models due to their efficiency in extracting water-bodies, and the different combinations were implemented to improve the extraction quality of water bodies and to reduce the calculation time.

After comparing our results with other current methods, we confirmed the effectiveness of our approach, we extracted the water bodies with a segmentation quality that exceeds 90%. We also reduced the calculation time by 80%.

Copyright
© 2025 The Author(s)
Open Access
Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

<Previous Article In Volume
Next Article In Volume>
Volume Title
Proceedings of the First International Conference on Artificial Intelligence, Smart Technologies and Communications (AISTC 2025)
Series
Advances in Intelligent Systems Research
Publication Date
5 August 2025
ISBN
978-94-6463-805-9
ISSN
1951-6851
DOI
10.2991/978-94-6463-805-9_15How to use a DOI?
Copyright
© 2025 The Author(s)
Open Access
Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Cite this article

risenwbib
TY  - CONF
AU  - Abdelali Benali
AU  - Hayet Kharbouch
AU  - Yesma Bendaha
PY  - 2025
DA  - 2025/08/05
TI  - New Architecture of Deep Learning using DeepLabV3+ ResNet50 and ResNet18 to Extract Water Bodies
BT  - Proceedings of the First International Conference on Artificial Intelligence, Smart Technologies and Communications (AISTC 2025)
PB  - Atlantis Press
SP  - 127
EP  - 140
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-805-9_15
DO  - 10.2991/978-94-6463-805-9_15
ID  - Benali2025
ER  -