Page 66 - Summer 2021

ECHO CLASSIFICATION
known. Because the generic function is not connected to the physical parameters of the object and environ- ment, related quantitative information cannot directly be extracted. However, if the function is applied under controlled conditions (such as with the same known types of scatterers and environment), then deviations in the parameters of the generic function can be used in classification and discrimination.
Sonar Applications
Assessing Fish Abundance
Fishes and the generally much smaller organisms, zooplank- ton, are prevalent throughout the world’s oceans. These two types of organisms are a major source of protein, both for larger organisms such as whales and for humans. Given the breadth of the importance of fishes and zooplankton, scientific echosounders are routinely used to assess their abundance and study their behavior (Stanton, 2012).
A common method used to estimate the abundance of fish and zooplankton is to relate the energy of the echoes from the echosounders to the number of organisms. However, it is commonly the case where an occasional large fish is feeding on a dense aggregation of zooplankton. In this case, the occasional echo from the fish can get washed away in the calculation of echo energy and therefore will not be counted. This issue can be addressed through use of echo statistics, where the presence of a dense aggregation of zooplankton forms the “background” or “baseline” echo
and the occasional large fish shows up as a “blip” above the background echoes (Stanton and Clay, 1986; Lee and Stanton, 2015). By using a statistical analysis, the two types of echoes can be separated. The fish echo occurs principally in the tail of the echo histogram, separate and distinct from the zooplankton echo (Figure 5). Through modeling of the echoes, the area under the tail can be used to count the fish.
Similar Applications: Submarines and Bats
The above example is conceptually similar to other types of applications: replace the individual fish with a sub- marine and replace the zooplankton with other various sources of scattering in the ocean such as the rough seafloor and sea surface (Figure 1). The echo from the submarine now creates a blip above the background echoes from the seafloor and sea surface. The process of detecting and classifying the submarine within this back- ground clutter is broadly similar to the fish/zooplankton problem. Here, the statistics of the echoes from the sub- marine and clutter must first be characterized separately and then be used to discriminate between each other.
The comparisons continue with terrestrial applications. For example, in the case of a bat echolocating on a moth flying in a forest, the moth creates the echo of interest and the trees and bushes are the background clutter.
Medical Ultrasound
The statistics of echoes from live tissue have been found to be useful in classifying echoes from different types
Figure 5. Sonar counting of occasional large fish feeding on a dense cloud of small zooplankton. Echoes from the zooplankton comprise most of the histogram (top right). The echoes from the occasional large fish appear in the “tail” of the histogram from which they can be counted. Adapted from Stanton, 1985.
66 Acoustics Today • Summer 2021