Abstract

Special-purpose diffractive optical devices are often designed by iterative methods without consideration of convergence properties and related factors that affect performance. We examine the properties of iterative algorithms in a vector-space setting and illustrate, with examples, differences in convergence performance based on starting point, sequential versus parallel projections, and intersecting versus nonintersecting sets.

© 1999 Optical Society of America

Application of the parallel generalized projection algorithm to the control of two finite-resolution deformable mirrors for scintillation compensation

Jeffrey D. Barchers
J. Opt. Soc. Am. A 19(1) 54-63 (2002)

Signal processing under uncertain conditions by parallel projections onto fuzzy sets

David Lyszyk and Joseph Shamir
J. Opt. Soc. Am. A 16(7) 1602-1611 (1999)

Processing halftone color images by vector space methods

Li Liu, Yongyi Yang, and Henry Stark
J. Opt. Soc. Am. A 23(2) 247-257 (2006)

Improved compensation of turbulence-induced amplitude and phase distortions by means of multiple near-field phase adjustments

Jeffrey D. Barchers and Brent L. Ellerbroek
J. Opt. Soc. Am. A 18(2) 399-411 (2001)

Restoration of signals from their signed Fourier-transform magnitude by the method of generalized projections

Eitan Yudilevich, Aharon Levi, G. J. Habetler, and Henry Stark
J. Opt. Soc. Am. A 4(1) 236-246 (1987)