Theoretical Aspects of Target Classification: Physical Optics and Radon Transform Methods,
Abstract
Target classification is set up as a multidimensional inverse scattering problem for electromagnetic waves. The target is considered to be of compact support residing in a medium with constant wave speed. The scattered field produced by a prescribed incident field produced by a prescribed incident field is represented by its equivalent sources in terms of Huygens' principle; for the sake of simplicity, inversion of this representation is outlined for a scalar quantity, say a scalar potential. It is illustrated that the introduction of the physical optics or weak scatterer approximation linearizes the inverse problem resulting immediately in an inversion formula for the characteristic or singular function of the scatterer if the scattered field is measured as function of frequency - or as broadband transients - on a closed surface sufficiently remote from the target in an either bistatic or monostatic experimental arrangement. When interpreted in the time domain the resulting far-field algorithms turn out to be of the backprojection type inversion of the Radon transform as it is a basic tool in computerized tomography. To get rid of the far-field approximation the wavefield backpropagation principle is utilized to define a generalized holographic field, which is explicitly related to the minimal energy components of the equivalent sources according to the Porter-Bojarski integral equation; integration with regard to frequency yields a unique explicit solution for the geometry of the target if and only if the physical optics or weak scatterer approximation holds.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1987
- Accession Number
- ADP005649
Entities
People
- K. J. Langenberg