Abstract

KAZAKOV, Vladimir  and  AFRIKANOV, Sviatoslav. Sampling - Reconstruction Procedure of Gaussian Fields. Comp. y Sist. [online]. 2006, vol.9, n.3, pp.227-242. ISSN 2007-9737.

The description of the optimal Sampling - Reconstruction Procedure (SRP) of Gaussian fields is given on the basis of the conditional mean rule when the quantity of samples is limited. The Gaussian fields are described by two types of space covariance function: exponential and Gaussian. A lot of both reconstruction and reconstruction error surfaces are obtained by numerical calculation. We changed the type of the covariance functions; the type of sampling (uniform: triangular, square, etc. and non - uniform: polar, spiral, and arbitrary); the quantity of the samples; the distances between the samples; and radii of the covariance functions of both axes. We demonstrate how all above mentioned factors influence on principal optimal SRP characteristics. The results of the calculations have clear interpretations.

Keywords : Gaussian Fields; Uniform and Non - Uniform Sampling; Reconstruction Functions; Reconstruction Error Functions.