Recent advances in deep learning (DL) and unmanned aerial vehicle (UAV) technologies have made it possible to monitor salt marshes more efficiently and precisely. However, studies have rarely compared the classification performance of DL with the pixel-based method for coastal wetland monitoring using UAV data. In particular, many studies have been conducted at the landscape level; however, little is known about the performance of species discrimination in very small patches and in mixed vegetation. We constructed a dataset based on UAV-RGB data and compared the performance of pixel-based and DL methods for five scenarios (combinations of annotation type and patch size) in the classification of salt marsh vegetation. Maximum likelihood, a pixel-based classification method, showed the lowest overall accuracy of 73%, whereas the U-Net classification method achieved over 90% accuracy in all classification scenarios. As expected, in a comparison of pixel-based and DL methods, the DL approach achieved the most accurate classification results. Unexpectedly, there was no significant difference in overall accuracy between the two annotation types and labeling data sizes in this study. However, when comparing the classification results in detail, we confirmed that polygon-type annotation was more effective for mixed-vegetation classification than the bounding-box type. Moreover, the smaller size of labeling data was more effective for detecting small vegetation patches. Our results suggest that a combination of UAV-RGB data and DL can facilitate the accurate mapping of coastal salt marsh vegetation at the local scale.

In the tidal flats of the Nakdong Estuary, eight weirs were installed as part of the Four Major River Restoration Project in 2011, and the environment changed from a flowing stream to a still water stream. As the Nakdong River’s weir was permanently opened in February 2022, the topography and ecological environment are expected to large change. In this study, Unmanned Aerial Vehicle (UAV) photogrammetry was conducted on the tidal flats of the Nakdong Estuary in November 2021, the environment before the Nakdong River floodgates were opened. The study area was surveyed using the Network-RTK (Real-Time Kinematic) method to obtain Ground Control Point (GCP), and using an UAV, orthographic image and digital elevation model were generated for an area of 3.47 ㎢ near Jin-u island and 2.75 ㎢ near Shin-ja island. A result of spatial resolution of 1.8 cm was obtained, the result was verified using checkpoints, and results with accuracy exceeding 1 cm were obtained in both Sin-u Island and Jin-woo Island. In the future, changes in the topography and sedimentation environment of this area are expected, so it will be useful data for various research and conservation management.

The Spartina anglica in the tidal flat at the southern part of Ganghwado, it is known that the distribution area has gradually expanded since it was officially announced as invasive alien species in 2015. The government and local governments are continuing their efforts to remove the S. anglica, and for this, continuous distribution change monitoring is required. This study extracted the data of distribution and extent area of S. anglica from Zenmuse P1 sensor, and generated the high-resolution Digital Elevation Model (DEM) from Zenmuse L1 sensor. Optical and Lidar images were photographed at an altitude of 70 m, and Ground Sampling Distance (GSD) of optical images was obtained at 0.9 cm and GSD of lidar images at 5 cm spatial resolution. However, the data are resampled and provided in GSD 25 cm to comply with the "National Spatial Information Security Management Regulations of the Ministry of Land, Infrastructure and Transport" and "Security Business Regulations of the National Intelligence Service".