Dimensional Images From a High-Resolution Underwater Acoustic Imager

Abstract

Abstract A underwater acoustic imaging system capable of working in extremely turbid conditions and producing high resolution dimensioned images has been developed. It is suitable for inspecting pipelines and offshore structures. A large development program has been carried out in Australia as a collaborative effort between a number of organisations to overcome the traditional problems with underwater imaging. The system uses megahertz frequency sound, sparse array technology, one-bit digitisation and chirp illumination. Acoustic three dimensional images with range and orthogonal dimensions have been obtained from a prototype underwater acoustic imaging system and examples are presented. Resolutions of millimeters have been obtained. This high resolution, large image (103x103x103 voxels) system is a new-to-the-world capability. This paper is a step in diffusing the new technology so that underwater acoustic imaging may become an innovation in offshore inspection and measurement. Introduction A new technology has been demonstrated for generating three dimensional near video quality images using underwater acoustics. There are two technologies for producing focused acoustic images, lenses and beam forming from twodimensional arrays. The lens has been demonstrated for two dimensional images by Belcher (1998) using a onedimensional array in the focal plane. The lens system places an array of acoustic sensors at the focal plane of the system and avoids the processing load of beam forming. Erikson (1997) discusses a more capable two-dimensional focal plane array. Beam formed images rely on the ability of the signals from an appropriately spaced array of acoustic sensors to be delayed in time and added to make the reflected pressure signals in phase for a particular voxel. For a large imaging volume this can become a very large processing load. The Underwater Acoustic Imaging technology is founded on five elements. Sparse arrays, one-bit digitisation, chirp ensonification, economical acoustic tiles and correlation flattening. While all of these elements have been used in some context before, it took the implementation of them all in a single system to produce a potential innovation. The sonar equation sets limits to performance that are determined by the source level and the system noise. While the source level is felt to be limited by the cavitation pressure, even before that magnitude the acoustic signal is nonlinear. Non linear acoustic signals distort during propagation and make replicate correction more difficult. System noise limits the range at which an object of given target strength can be seen above this noise. Arrays do not focus exactly on a single voxel but allow acoustic energy from adjacent voxels to leak into the picture. Replicate correlation of the received signals with the chirp ensonification, which provides focusing in the range direction, does the same. The clutter from elements of the target limits the size of objects that can be viewed clearly by restricting the contrast. Technology For an active acoustic imaging system such as the present, the acoustic pressure on the face of the receive array depends on the source level, the range of the target, its target strength and the absorption and scattering of water.