Imaging algorithms for a strip-map synthetic aperture sonar: Minimizing the effects of aperture errors and aperture undersampling

Imaging the sea floor using high-precision synthetic aperture sonar (SAS) techniques is now at the stage where the efficiency and the robustness of the various imaging algorithms are of concern. There have been several block processing algorithms developed for relatively narrow-band-, narrow swath-, and narrow beamwidth synthetic aperture systems mainly for use by the synthetic aperture radar (SAR) community. These algorithms are summarized and their relevance to the slower speed of propagation, broad-band, broad swath-, and broad beam-width sonar equivalents are noted. Additional algorithms intended to ameliorate distortions infected by motion errors of the tow fish and medium turbulence are also assessed. One of the significant differences between the sonar and radar systems is that most synthetic aperture sonars travel faster than that required to meet the spatial sampling criterion and so the aperture is under- or insufficiently sampled. The digital spotlighting approach can be shown to reduce the grating-lobe images generated by this undersampling to a significant degree. The operational effectiveness of these various algorithms are shown on real data as collected by an ocean-going, boat-towed, rather than a rail or otherwise guided, sonar. What is important is that these algorithms in various combinations can ultimately produce near diffraction-limited imaging on real data. Typical results are shown when using the Kiwi-SAS to image point retro-reflectors (either as isolated targets or deployed in pairs) on a sea floor of bland silt. To date, no unclassified SAR or SAS uses the range or along-track spatial bandwidths employed by the Kiwi-SAS. The final SAS image resolution of 16 cm x 5 cm is a considerablely finer resolution than achieved by any SAR of equivalent carrier wavelength. The fine resolution is due to the correspondingly high spatial bandwidths covered by the system; that of range due to the chirp bandwidth coupled with the slow speed of sound in water and that of along-track due to the small real apertures employed. Access to this wide spatial bandwidth makes the applicability of normal SAR algorithms uncertain and we explore some of the trade-offs.