Automated Detection Of Algal Blooms In Satellite Images Using K-Means Clustering With NDVI And NDWI Indices

Harmful algal blooms (HABs) greatly affect water bodies, aquatic ecosystems, water quality, and human health. The present work suggests an automatic algal bloom detection approach from satellite imagery by incorporating k-means clustering along with the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI). These indices are used to increase feature extraction for accurate segmentation of water areas, vegetation, and algal bloom areas. The k-means clustering algorithm was applied with k=4. NDVI and NDWI values were normalized between -1 to 1, effectively differentiating bloom-affected areas, with an average NDVI of 0.38 for vegetation and an average NDWI of 0.12 for water bodies. The suggested method enables real-time monitoring of water quality, advancing sustainable management of ecosystems.