Estimation of single-beam echo sounder signal for riverbed classification

Newly defined are two parameters which may characterize the sediment types as well as roughness index (E₁) and hardness index (E₂). Then, a method using BP neural network, which includes four parameters as an input layer vector, is proposed to accurately classify the riverbed sediments which have rough floor and various types of sediments. The data obtained from Single-Beam Echo Sounder whose operation frequency is 200 kHz and beam angle is 10°, and true ground samples are used and the method using only two parameters (E1 and E2) and the proposed method are compared to verify the effectiveness of the proposed method. The proposed method results with an accuracy of 84.95% on rough riverbeds.