Coherent and semi-coherent processing of limited-aperture circular synthetic aperture (CSAS) data

Circular synthetic aperture sonar (CSAS) traditionally involves the coherent processing of 360 degree scattering information from acoustic targets. To obtain 360 degree scattering information, a source may circle around a central target field and constantly illuminate targets from multiple aspects. Another method of obtaining CSAS data is to fix the source location and spin a target on a rotating mount. Following data reception, a variety of methods in the Fourier or time-domain may be used to construct images. For certain targets, resonances and elastic effects can interfere with the specular portions of backscattered echoes. The time-delay associated with elastic or resonant responses destroys the uniqueness of the location to which the signal is mapped, and occasionally these resonant features can be mapped directly on top of target specular features, causing destructive interference and reduced image clarity. Destructive interference can be reduced and image clarity enhanced by incoherently summing separate images generated from sub-apertures of CSAS data. Additionally, limiting the aperture and frequency band of the pre-processed data before applying an imaging algorithm is an effective method for understanding and localizing various elastic and non-elastic target responses. In a solid 3-to-1 cylinder, for example, effects such as meridional and face-crossing rays cause well defined image features that are prominently visible when limiting the aperture to the angular portions in which these rays are the dominant elastic effects. Further analysis may be obtained by masking portions of these sub-aperture images and reversing the imaging process. This can be used to directly relate target image features to the angular frequency response (colorplot) of the target. An added benefit of this reversal process is that signals from surrounding objects, and the random noise spread throughout the image scene can be rejected by an image masking process, and the resulting time-domain information has an enhanced signal-to-noise ratio. This effect has been successfully demonstrated on data acquired in field-tests, and in controlled laboratory experiments with real and replicated Unexploded Ordnance (UXO) objects. A "Projection-Slice" based CSAS script has been tested on data acquired at sea by an unmanned vehicle, as well as in laboratory experiments from UXO objects placed on a rotational mount. The laboratory based full scale UXO datasets were acquired in a control