Optimization of oblique drone photogrammetry for avoiding sun glint in submerged seagrass mapping

Abstract

Submerged vegetation, such as seagrass meadows, plays a pivotal role in blue carbon sequestration and provides synergistic benefits to ecosystem services. Drone-based monitoring of blue carbon in submerged seagrass meadows is a cost-effective and high-resolution approach. However, the presence of sun glint on the water surface, particularly under sunny conditions, hampers the effectiveness of drone-based mapping. This study presents an improved drone-based oblique photogrammetry method for submerged seagrass meadows mapping. This method is specifically designed to avoid sun glint by leveraging the derived theory that correlates solar angles with the drone gimbal pitch angle. Field experiments conducted in the Futtsu Tidal Flat of Tokyo Bay successfully investigated the feasibility of the proposed method. The results indicate that it is crucial to maintain a gimbal pitch angle of less than 55° to obtain clear orthophotos when mapping submerged seagrass meadows. Variations in the solar azimuth angle significantly constrained the effective mapping window for oblique photogrammetry under sunny conditions. Compared to vertical photogrammetry, oblique photogrammetry exhibited a 10.9% higher reprojection error, an 8.9% increase in ground sample distance, and a 24.2% reduction in automatic tie points per image. This method extends the effective mapping window by 40 minutes under sunny conditions with controlled accuracy loss (8.9% GSD increase), demonstrating operational value for blue carbon quantification in submerged seagrass ecosystems. • By optimizing the relationship between gimbal pitch angle and solar angles, sun glint is effectively avoided, thereby extending the mapping window. • Maintaining a gimbal pitch angle below 55° is essential for obtaining clear orthophotos. • Vertical and oblique photogrammetry yield only slight differences in orthophoto quality.

Topics & Concepts