Raymond Lim
Code 130B, Coastal Systems Station, Panama City, FL 32407-7001
Conventional sonar systems for bottom searching underwater coastal environments have limited long-range classification capabilities. This is especially true when objects searched for are significantly buried so that imaging is difficult. To investigate the feasibility of nonimaging methods of classification (e.g., based on isolating resonances in the echo from the desired object), models that faithfully predict the acoustic response of known objects in various states of burial are required. Because the response of the object can be strongly modified by the structure of its local environment, an appropriate model must account for effects due to propagating the acoustic field to and from the object, the scattering itself, and the interaction of the object with its local environment. The present talk will discuss how these issues can be handled in an exact fashion via a transition-matrix formulation both for fully buried and partially buried objects. In either case, the solution involves identifying a suitable set of global basis functions. [Work supported by ONR and the CSS IR program.]