Monitoring Restored Mangroves in Porto do Mangue, Brazil: Assessing RGB Vegetation Indices and DEM from Drone Imagery

Abstract

This study evaluated the use of RGB images obtained by Remotely Piloted Aircraft (RPA) for monitoring mangroves undergoing environmental recovery, located in Porto do Mangue-RN, Brazil. Six spectral indices (VARI, GLI, MGVRI, MPRI, RGVBI, and ExG) were applied with the aim of classifying the areas into three thematic classes: exposed soil, low vegetation, and shrub vegetation. The image processing allowed the generation of orthophotos and Digital Elevation Models (DEM), which provided support for validating the obtained results. The spectral indices were individually tested using QGIS software and analyzed in comparison with field data and the DEM. The Excess Green Index (ExG) showed the highest accuracy and consistency in identifying the classes, with emphasis on shrub vegetation, achieving values very close to the reference area. The other indices showed limitations, especially in scenarios with high spectral variability and the presence of shadows and wet soil. The use of high-resolution RGB imagery associated with simple spectral indices proved to be an accessible and efficient tool for monitoring mangrove areas under regeneration, with potential to support environmental recovery actions, reforestation, and public preservation policies. The adopted methodology provides support for environmental management in contexts of limited resources and reinforces the role of RPAs in monitoring sensitive ecosystems.