Acoustic Classification with Single-Beam Echosounders: Processing Methods and Theory for Isolating Effects of the Seabed on Echoes

Analyzing the details of echoes from single-beam sounders is well established as a basis for seabed classification. Both echo shape and amplitude are analyzed to make maps of seabed sediment classes. For accurate acoustic classification of the seabed, the echoes should be influenced only by properties of the seabed. This paper surveys methods to remove artifacts caused by water depth, system parameters and seabed slope, while discussing processing methods for seabed classification. High-speed data acquisition and depth and slope compensation is discussed. Field data and BORIS simulations are presented. Potential seabed classification or characterisation processes include: frequency shifts, relative strengths and shape of the coherent echo, and residual echo durations. For seabeds with any significant roughness, the echo is not at the same frequency as the transmission. These frequency shifts may be valuable for characterisation. Classification by segmentation requires depth, slope and system independent digital versions of the echo envelopes. Because echo power can be distributed anywhere within the pulse bandwidth, the best acquisition method is to sample fast enough to capture the carrier, filter, and form envelopes with the Hilbert transform. This allows direct estimation of the coherent echo by full wave-form stacking. The coherent echo has great potential for seabed characterization.