Fine Resolution Imaging of Space Objects.
Abstract
This report describes the results of a two-year research effort to investigate the reconstruction of fine-resolution images of space objects using earth-bound optical telescopes despite the turbulence of the atmosphere. The results of this research are an indication that, using an iterative reconstruction algorithm, it is feasible to reconstruct diffraction-limited images from the Fourier modulus (or autocorrelation) data provided by stellar speckle interferometry. Methods of compensating for systematic errors and noise present in astronomical data were developed and were applied to binary star data, from which a diffraction-limited image was successfully reconstructed. It was analytically shown that the following types of objects are uniquely specified by their Fourier modulus: objects consisting of separated parts satisfying certain disconnection conditions, objects consisting of collections of distinct points, and objects with radial symmetry. Experimental indications are that complicated two-dimensional objects are usually unique. An equivalence of the error-reduction iterative reconstruction algorithm and the steepest-descent gradient search method was shown, and the convergence of this algorithm was proven. However, the input-output iterative algorithm and other gradient search algorithms were found to converge much faster in practice. A survey of other applications of the iterative algorithm was performed, and it was found to be remarkably useful in solving problems for a wide range of applications.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1982
- Accession Number
- ADA113911
Entities
People
- James Fienup
Organizations
- Environmental Research Institute of Michigan