LAUSR

The accurate and efficient detection of the water depth of coral reefs through remote sensing imagery is crucial for navigation, marine engineering, and marine safety, among others. This study proposed a water depth inversion method that is processed using a two-step strategy. First, a log-dual-band ratio transform model is employed by combining blue, yellow, red, and red edge bands. Second, a multivariable linear regression model is constructed to determine water depth. The experimental study is conducted in a reef island in the South China Sea, and the WorldView-2 satellite data is used as the test data. Results demonstrated that the proposed method needs a few parameters in the model, and can effectively and reliably characterize the correlation between the spectral features and water depth.