Image restoration by the method of least squares.
Abstract
A method based on generalized least squares is applied to the reduction of speckle interferometric data. A model for the spread function is assumed, and the free parameters of the spread function are fitted to the data. The object profile is formulated as a function that depends on the chosen unknowns. The unknowns may be star radius, or the mesh points of the object. The simplest model for the spread function is the product of an infinite twodimensional Gaussian process and a weight function given by the Gaussian profile of the long exposure point spread function. The model can be written in either the spatial domain or as a Fourier series. Isoplanicity of the object is not assumed. All known effects of aberration and uncorrected atmospheric dispersion can be taken into account in the model equation. It is believed that the limiting magnitude of the method is at least as good as the theoretical limit.
 Publication:

Astronomy and Astrophysics
 Pub Date:
 December 1978
 Bibcode:
 1978A&A....70..777V
 Keywords:

 Image Processing;
 Imaging Techniques;
 Interferometry;
 Least Squares Method;
 Speckle Patterns;
 Telescopes;
 Atmospheric Effects;
 Autocorrelation;
 Computer Programs;
 Data Reduction;
 Matrices (Mathematics);
 Astronomy;
 Image Processing:Speckle Interferometry