3-D acoustic imaging of broadband SAS data

The difficult conditions encountered in the littoral region require development of flexible sonar processing technologies. One technical approach to increase the flexibility of active sonar has been the development of broadband sonar for multifrequency analysis. Physics based target models of acoustic backscatter use frequency as a prime variable, often defined in terms of radius, a, and acoustic wavenumber, k. For example, fluid filled spheres insonified at low frequencies exhibit sharp peaks and nulls in the acoustic backscatter. As ka increases marine biologists have predicted and observed a frequency variation in target strength associated with shape in the acoustic return of biologics that conform to fluid filled and elastic object models. The broadband synthetic aperture sonar (BBSAS) at CSS enables one to add another dimension (frequency) to the target feature space without compromising resolution. The processing enables a user to construct a 3-D image description around targets of interest. A modified STFT, Choi-Williams, and Binomial transformations, are compared for the frequency decomposition stage of the beamformer. A trade-off between processing complexity and preservation of the target characteristics is discussed.