Ghosh et al. (2026) A Hybrid Approach for Delineation of Water Bodies Using Filtering and Mathematical Operations on Sentinel-2 Satellite Images

Identification

• Journal: Lecture notes in networks and systems
• Year: 2026
• Date: 2026-01-01
• Authors: Deblina Ghosh, Shivani Saxena
• DOI: 10.1007/978-3-032-12827-0_35

Research Groups

• Department of Mathematics, Banasthali Vidyapith, Banasthali, Rajasthan, India
• Department of Humanities and Applied Science, KCCEMSR, Thane, Maharashtra, India
• Department of Physical Sciences, Banasthali Vidyapith, Banasthali, Rajasthan, India

Short Summary

This paper proposes a generic hybrid image processing approach combining filtering techniques and mathematical morphological operations to accurately delineate water bodies from Sentinel-2 satellite images, demonstrating significantly improved performance over conventional methods.

Objective

• To propose and evaluate a generic, fast, and lightweight hybrid approach for delineating water bodies from Sentinel-2 satellite images using a combination of image filtering techniques and mathematical morphological operations.

Study Configuration

• Spatial Scale: Delineation of water bodies from Sentinel-2 satellite images; tested on a simulation of 36 water body images.
• Temporal Scale: Not explicitly defined for the data used; the method is designed for static image analysis.

Methodology and Data

• Models used: Bilateral filter, Median filter, CLAHE filter, Tophat Transform, Hysteresis Thresholding, Dilation, Morphological Reconstruction.
• Data sources: Sentinel-2 satellite images (specifically, a simulation dataset of 36 water body images).

Main Results

• The proposed hybrid method achieved an average Accuracy of 89.57%, Precision of 90.28%, Recall of 81.43%, and an F1 score of 0.85.
• In comparison, conventional filtering techniques reported an Accuracy of 62.57%, Precision of 55.22%, Recall of 83.61%, and an F1 score of 0.61.
• The proposed method offers fast computation, a lightweight architecture, and is effectively applicable to a wide range of satellite images, ensuring reliable and robust water body detection.

Contributions

• Introduction of a novel hybrid approach for water body delineation that synergistically combines specific image filtering techniques (Bilateral, Median, CLAHE) with a sequence of mathematical morphological operations (Tophat Transform, Hysteresis Thresholding, Dilation, Morphological Reconstruction).
• Demonstrated significant quantitative improvements in delineation accuracy (89.57% vs 62.57%), precision (90.28% vs 55.22%), and F1 score (0.85 vs 0.61) compared to conventional filtering methods.
• The developed method is characterized by its generic applicability, computational efficiency, lightweight architecture, and robustness across diverse satellite imagery.

Funding

• Not specified in the provided paper text.

Citation

@article{Ghosh2026Hybrid,
  author = {Ghosh, Deblina and Saxena, Shivani},
  title = {A Hybrid Approach for Delineation of Water Bodies Using Filtering and Mathematical Operations on Sentinel-2 Satellite Images},
  journal = {Lecture notes in networks and systems},
  year = {2026},
  doi = {10.1007/978-3-032-12827-0_35},
  url = {https://doi.org/10.1007/978-3-032-12827-0_35}
}