Imaging of buried objects by low frequency SAS

Abstract

Search for buried objects are normally done by normal incidence sonar, working with wide-band low frequency signals. Especially the parametric sonar has proven to be a reliable tool to detect buried objects, as well as for mapping of internal stratification in sub-bottom layers. However, due to the small foot-print searching with normal incidence sonar is tedious. Also, in a mine hunting situation, the platform has to pass over the mine, which of course is undesirable. Forward looking and side-scan sonar have the capability to detect and classify objects at some distance from the platform. Their azimuth resolution is dependent on aperture length compared to the wavelength. A long aperture gives better resolution than a short. For proud or floating targets it is always possible to increase resolution by increasing frequency although that would decreases the possible operating range. To find buried objects the frequency has to be low for the sound to penetrate into the bottom. Synthetic Aperture Sonar (SAS) is a signal processing method designed to increase the resolution without changing the frequency. The physical aperture is moved and data from several positions are used for synthesizing a longer synthetic aperture. The long wavelengths used for buried targets means that the aperture has to be increased even more to achieve good resolution. Obviously SAS is even more useful for bottom penetrating sonar than the extensively reported purpose of imaging objects lying on the seafloor. We have developed a side-scan sonar using SAS processing to image buried objects. The sonar has been tested at our pontoon laboratory, which has been equipped with a 16 m long rail for SAS related research. Objects of various types have been placed both on the bottom and in the sediment. A 20 - 30 kHz frequency modulated signal was transmitted from a small transducer and echoes were received by a 1.5 m horizontal linear array. SAS images of both proud and buried targets will be shown indicating that SAS is an obvious choice for future mine-hunting sonar, both penetrating and none penetrating.