Extracting sonar relative beam patterns for multi-sector multibeam sonar

Abstract

The use of multibeam acoustic backscatter data for bottom characterization is currently being attempted by many researchers to aid geological, biological, and engineering projects. Ideally the absolute bottom backscatter strength would be measured, but in reality the reported data are overprinted by system-related geometric and radiometric effects. In real time, manufacturer-applied gain only partly reduces these effects. Existing post-processing algorithms undertake improved but still imperfect corrections to better account for these residual artifacts. The geometric effects include changing range, grazing angle and insonified area across the swath, whereas the radiometric effects include the angular variation in the transmitted energy and the receiver sensitivity. Recent developments in motion stabilization that involves multiple sectors, which are used to achieve higher and more equal sounding density, have added significantly more radiometric complications to the backscatter imagery. Before the backscatter data can be used for classification, either in the form of a mosaic or in the form of backscatter strength angular response curves, these remaining artifacts in the data have to be properly minimized. The residual artifacts reflect the fact that existing empirical beam pattern corrections imperfectly account for geometry and radiometry, and do not adequately distinguish between grazing and sonar relative angle. This research develops a new method of reducing the backscatter data by explicitly differentiating between seafloor angular response and radiometric artifacts. The new method further differentiates between along-track and across-track radiometric beam patterns. The developed method does not require any prior knowledge of seafloor characteristics. It is capable of propagating standard deviation from the backscatter data to the extracted radiometric beam pattern. This enables the user to access the reliability of extracted radiometric beam patterns.