Reversible line‐scan and circular synthetic aperture sonar processing with applications to object classification.

Abstract

Spectral response as a function of aspect, often called “acoustic color,” is sometimes used to classify objects. The ability to utilize reversible imaging algorithms to enhance spectral features and to isolate the acoustic color plots of adjacent targets were previously demonstrated in the line‐scan case [Marston et al., J. Acoust. Soc. Am. 128, 2461 (2010)]. Sometimes the resonant or ringing part of an object’s response is isolated from the spatially localized background. This was verified by experiments at Washington State University with a metal cylinder where the ringing part was modeled using quantitative ray theory. An overview of multiple inversion techniques and the importance of the assumptions behind the beam forming processes involved will be given. Current research additionally concerns the application of reversible circular synthetic aperture sonar algorithms to enhance our understanding of the relationship between CSAS images of objects and particular spectral features present in color plots. An imaging approach based on the Fourier‐slice theorem, and the subsequent reversal process will be given, along with examples of the reversal process being used to associate object features with specific spectral features. [Work supported by ONR.]