Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Speckle-image, phase reconstruction techniques compared

Open Access Open Access

Abstract

The Knox-Thompson (KT), triple-correlation (TC), and phase-gradient (PG) algorithms have been shown capable of reconstructing the phase of stellar spatial spectra upto the diffraction limit of large telescopes. Combined with the modulus measurements of Labeyrie’s speckle interferometry, these techniques offer from 10 to 50 times improvement in image resolution over the 2-sec of arc limit imposed on conventional astronomical observations by atmospheric turbulence. We report results of a comparative study in which the performance of the three techniques is evaluated with a common set of photon-limited speckle-image data from both simulated and real astronomical sources. The regime of interest is < 1 photon/ speckle and <500 photons/image. In these conditions PG gives excellent reconstructed images and is the most computationally efficient of the three. KT gives slightly poorer results in severe atmospheric conditions, can be improved by image centroiding, but experiences a further distorting effect at very low photon levels. TC requires the greatest computational effort but appears to be more robust with respect to certain types of data defect. Extension of TC can give a small improvement in image quality but at a high computational cost.

© 1988 Optical Society of America

PDF Article
More Like This
Photon-Limited Performance of Three Stellar Speckle Image Reconstruction Algorithms

G.J.M. Aitken, R. Johnson, and J. Meng
TuA3 Quantum-Limited Imaging and Image Processing (QLIP) 1989

Photon-limited phase-gradient stellar image reconstruction

G. J. M. Aitken and R. Johnson
MO4 OSA Annual Meeting (FIO) 1987

Phase-Diverse Speckle Interferometry

Richard G. Paxman, Timothy J. Schulz, and James R. Fienup
TuA2 Signal Recovery and Synthesis (SRS) 1992

Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.