Assessment of UAV-Based Structure-from-Motion Photogrammetry for Detecting Coastal Benthic Habitats and Comparison with Multibeam Bathymetry

Coastal habitats are essential to human life by providing oxygen, sequestering carbon, conserving biodiversity, and providing a wide range of ecosystem services that contribute to human well-being. Determining the current status and managing marine habitats, such as mangroves, seagrass meadows, and coral reefs, across various scenarios is of paramount importance. While commonly used hydrographic surveying techniques, like acoustic techniques, are available for this purpose, recently, Unmanned Aerial Vehicle (UAV)based systems have been used owing to their costeffectiveness, rapidness, and flexibility. This study aims to assess the UAV-based Structure-from-Motion (SfM) photogrammetry for the detection of coastal benthic habitats. In this regard, Mean-Shift Segmentation and a supervised classification method based on the Support Vector Machine (SVM) algorithm were used. The classification results achieved high accuracy, with F1 scores >95%, demonstrating that the method used is quite effective and feasible for identifying benthic habitats in shallow waters. Moreover, the orthophoto was also cross-checked with a bathymetric model generated using a Multibeam Echosounder (MBES) in the study area. The study results showed that UAV technology can be effectively adopted for habitat mapping in shallow areas where water conditions are favorable, and its integration with bathymetric data can improve the overall analysis quality.