Integrated image segmentation techniques for high-resolution coastal habitat mapping: Advancing remote sensing for coastal ecosystem assessment

Abstract

This study presents an integrated approach to high-resolution coastal habitat mapping using advanced image segmentation techniques and remote sensing indices for the Torres Coastal Region, Brazil. The research leverages multiple spectral indices to evaluate and predict coastal ecosystem health, including the NDMI, BSI, FAI, NDBI, MNDWI, NDWI, NDRE, GNDVI, SAVI, and NDVI. The study employs the InVEST model to assess the Habitat Quality Index (HQI) for 2024 and predict changes in habitat quality by 2034. The findings reveal significant spatial variations in habitat quality, categorized into five classes: Very Low (208.76 sq. km), Low (105.25 sq. km), Moderate (196.85 sq. km), High (60.53 sq. km), and Very High (105.14 sq. km), over a total area of 676.53 sq. km. These classifications reveal insights into the spatial distribution of ecosystems and their varying resilience to anthropogenic pressures, such as urbanization, pollution, and climate change impacts. The research emphasizes the role of remote sensing indices, which provide a holistic view of environmental conditions, including vegetation health, water quality, soil properties, and land use patterns, contributing to the overall understanding of coastal ecosystem dynamics. The predicted Habitat Quality Index for 2034 offers crucial information for forecasting future environmental conditions and supporting evidence-based decision-making for coastal management. Furthermore, this approach can act as a valuable perfect for similar coastal areas, helping establish guidelines for integrating remote sensing tools in ecosystem health assessments globally.