Abstract

In many imaging applications, the wavelength of the illuminating wave is comparable in scale to the structural features of interest. Hence, the scattering and diffraction effects have to be incorporated in the inversion algorithms for the conventional diffraction tomography, in which the distribution of scattering centers can be recovered only if the scatters are weak or are slowly fluctuating. An inversion algorithm, which extends the limitations of conventional diffraction tomographic techniques based on the firstorder Born or the Rytov approximations is presented. This algorithm preserves a Fourier relationship between scattered-field data and the scatterer. When prior knowledge about the scatterer is available, the distorted-wave Born approximation can be used. This technique improves both the weak-scattering approximation and the required Fourier data interpolation and extrapolation. The procedure is illustrated with both simulated and microwave-scattered data taken at 10 GHz.

© 1990 Optical Society of America