Synthetic aperture sonar systems are employed to detect unexploded ordnance in marine environments. These systems use broadband, low frequency (below 50 kHz) waveforms to penetrate the sediment. Three-dimensional imagery is then reconstructed using delay and sum beamforming. Despite the complicated nature of these sonar systems, accurate detection, localization, and classification of objects remains challenging. The reconstruction algorithms, however, do not currently exploit all the available information in the signals. Recent work in biomedical ultrasound imaging has shown how differences in spatial coherence between targets and the background can be used to improve image quality. While the signals from a munitions survey system are expected to have similar properties, differences in array geometries and issues like uncompensated platform motion pose challenges to their adoption. This presentation will describe recent investigations applying coherence-based reconstruction to synthetic aperture sonar systems. By adapting algorithms from biomedical ultrasound such as delay-multiply-and-sum, as well as exploiting the multi-static nature of these systems, it is possible to recover vital clues about an object’s shape, pose, and burial depth.
Thomas E. Blanford (Tue,) studied this question.