Application of RGB UAV imagery to sea surface suspended sediment concentration monitoring in coastal construction site

Monitoring the spatial distribution of suspended sediment concentration (SSC) is of paramount importance, particularly in coastal construction sites where elevated SSC levels are inevitable. Remote sensing techniques using unmanned aerial vehicles (UAVs) have proven to be effective in estimating SSC distribution, making use of the optical spectral variations induced by SSC. The SSC-sensitive spectral range includes the red, green, and blue bands. This study demonstrates the feasibility of utilizing RGB images for SSC monitoring. The process begins with the acquisition of RGB images by a UAV, followed by ortho-rectification using the Structure-from-Motion Multi-View Stereo (SfM-MVS) algorithm. Subsequent steps involve gamma correction of the standard RGB image and radiometric correction, leading to the creation of a reflectance map. The spectral analysis establishes a clear relationship between SSC and the red band reflectance. As for the field scale application, we analyzed RGB images obtained from a construction site in South Korea. The spectral models for SSC were evaluated by varying the functional form, considering both linear-log and power-law models along with spectral bands and their ratios. As a result, it was revealed that the power law formula with the red band reflectance showed the highest correlation with SSC. As a result, nine snapshots of SSC distribution were reliably generated. This study underscores the practicality of employing a UAV-RGB camera for high-resolution SSC distribution monitoring, which is both cost-effective and efficient.