Abstract

An information divergence, such as Shannon mutual information, measures the distance between two probability-density functions (or images). A wide class of such measures, called α divergences, with desirable properties such as convexity over all space, was defined by Amari. Rényi’s information D_α is an α divergence. Because of its convexity property, the minimum of D_α is easily attained. Minimization accomplishes minimum distance (maximum resemblance) between an unknown image and a known reference image. Such a biasing effect permits complex images, such as occur in inverse synthetic-aperture-radar imaging, to be well reconstructed. The algorithm permits complex amplitudes to replace the probabilities in the Rényi form. The bias image may be constructed as a smooth version of the linear, Fourier reconstruction of the data. Examples on simulated complex image data with and without noise indicate that the Rényi reconstruction approach permits superresolution in low-noise cases and higher fidelity than ordinary, linear reconstructions in higher-noise cases.

© 1995 Optical Society of America

Bayesian cross-entropy reconstruction of complex images

B. Roy Frieden and Anisa T. Bajkova
Appl. Opt. 33(2) 219-226 (1994)

Minimum entropy–neural network approach to turbulent-image reconstruction

Yuanmei Wang and B. Roy Frieden
Appl. Opt. 34(26) 5938-5944 (1995)

Reconstruction of multidirectional interferometric data using an isoparametric finite-element method

Yan Zhang and Gary A. Ruff
Appl. Opt. 34(28) 6361-6372 (1995)

Discrimination of isotrigon textures using the Rényi entropy of Allan variances

Salvador Gabarda and Gabriel Cristóbal
J. Opt. Soc. Am. A 25(9) 2309-2319 (2008)

Image recovery by minimum discrimination from a template

B. Roy Frieden
Appl. Opt. 28(6) 1235-1243 (1989)