DEFECT IMAGING METHODS WITH SUPERRESOLUTION IN MULTIFREQUENCY DIGITALACOUSTIC HOLOGRAMS (REVIEW)

Image resolution is determined by the filtering properties of the inst rumentation of the acoustic imaging system, i.e., by the passband for temporal and spatial harmonics. The ray (longitudinal) resolution is d etermined by the frequency interval in which the holograms are compute d. Scatterer images can be obtained with longitudinal superresolution (resolution exceeding the Rayleigh limit) by using either an iterative method of adaptive extrapolation (modified Gershberg-Papoulis algorit hm) or an autoregression model to extrapolate the echo-signal spectra. In this article the adaptive extrapolation method is generalized to t he two-dimensional case; it can then be used in the stage of hologram reconstruction by projection in spectral space. This approach permits scatterers to be imaged both with ray resolution and with wavefront su perresolution. The efficiency of the investigated methods is tested in experiments on the reconstruction of images of models of point and ex tended inhomogeneities. The ray resolution is improved twofold in adap tive extrapolation and fourfold in application to the autoregression m odel of the echo-signal spectrum. The wavefront and ray resolutions ar e enhanced fourfold and twofold, respectively, in experiments on image processing by two-dimensional adaptive extrapolation. The application of different extrapolation methods in different stages of the imaging process improves the reliability of the results.