Improving riverbed sediment classification using backscatter and depth residual features of multi-beam echo-sounder...

Area of Science:

• Geophysics
• Acoustic Remote Sensing

Background:

• Riverbed and seafloor sediment classification is crucial for understanding aquatic environments.
• Acoustic remote sensing offers cost-effective, high-coverage solutions for sediment mapping.

Purpose of the Study:

• To classify riverbed sediment using multi-beam echo-sounder data and an empirical method.
• To evaluate the effectiveness of backscatter strength and depth residual features for sediment classification.

Main Methods:

• Empirical classification method using multi-beam echo-sounder data.
• Principal Component Analysis (PCA) to identify informative features.
• Clustering applied to principal components for sediment class assignment.

Main Results:

• Backscatter strength features effectively discriminate sediment properties.
• Depth residual features distinguish riverbed forms like riprap structures and ripples.
• Combined features provide complementary information on both sediment composition and riverbed structure.

Conclusions:

• The empirical method successfully classifies riverbed sediment.
• Combining backscatter and depth residual features enhances classification accuracy.
• This approach offers valuable insights into riverbed composition and morphology.