Comparative Evaluation of Water Indices for Water Body Mapping from Sentinel 2 Imagery: A Thresholding and SVM-Based Accuracy Approach

Author(s)	Mr. Shriram P. Kathar, Prof. Dr. Prapti D. Deshmukh
Country	India
Abstract	This study compares water body extraction techniques using Sentinel-2 images. It combines spectral index-based thresholding with a supervised machine learning approach. Four common water indices—NDWI, MNDWI, AWEI, and SWM—were analyzed using Otsu and Minimum thresholding methods. A Support Vector Machine (SVM) classifier assessed accuracy. The results show that the SVM classifier performed the best, achieving an overall accuracy of 99.79%, a Kappa coefficient of 0.9912, and an F1-score of 0.9901. This demonstrates outstanding precision and reliability. In contrast, thresholding methods were less effective, especially for indices like AWEI_nsh and AWEI_sh, revealing their sensitivity to data changes. The findings stress that while spectral indices effectively highlight water features, combining them with machine learning greatly enhances extraction accuracy. The study concludes that using Sentinel-2 images alongside SVM classification provides a strong method for accurate and efficient water body mapping.
Keywords	NDWI, MNDWI, AWEI, SWM, SVM
Published In	Volume 16, Issue 4, October-December 2025
Published On	2025-12-01

About IJSAT Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us	Message on WhatsApp	+91-9687-182-185	editor@ijsat.org

International Journal on Science and Technology