Shadow enhancement in synthetic aperture sonar imagery for improved target classification

The acoustic shadow of a target is a strong classification clue in naval mine hunting. A shadow of bottom targets appears at relatively short ranges (≤ about 200 m) in high resolution images from sidescan or synthetic aperture sonar (SAS), against a background of sea floor reverberation. In SAS however, definition of the shadow reduces with increasing synthetic aperture length, which is a measure of image resolution. This effect is similar to an optical shadow that is diffuse when an extended light source is used, but sharp with a point source. The longer the light source (the optical equivalent of the synthetic aperture), the more blurred the shadow becomes. This paper presents a SAS technique that enhances the shadow from a detected target by compensating blurring of the shadow in the image forming process. The proposed method intends to get the best of both worlds: good SAS resolution and good shadow definition. The method is tested on the basis of simulations and real data. It is conluded that automatic shadow contour detection is improved