Photogrammetry-Based Photic Seafloor Surveying and Analysis with Low-Cost Autonomous Underwater and Surface Vehicles

抄録

This study explores advanced methods for underwater visual surveys in the photic zone using low-cost autonomous underwater and surface vehicles to enhance marine ecosystem monitoring and analysis. It addresses the challenges of underwater photogrammetry, including data post-processing and analysis, by introducing geo-reference estimation techniques for autonomous-under-vehicle-mounted cameras. Through a novel validation method based on trajectory overlaps and application to real-world datasets, the research demonstrates the effectiveness of these approaches in complex underwater environments. The findings contribute to improving the accuracy of three-dimensional reconstructions of the seafloor, offering significant implications for marine conservation and ecosystem management.