Blind deconvolution post-processing of images corrected by adaptive optics

1995 ◽  
Author(s):  
Julian C. Christou
Keyword(s):  
Adaptive Optics ◽  
Blind Deconvolution ◽  
Post Processing

scholarly journals Blind deconvolution in astronomy with adaptive optics: the parametric marginal approach

2020 ◽  
Vol 496 (4) ◽  
pp. 4209-4220 ◽  
Author(s):  
R J-L Fétick ◽  
L M Mugnier ◽  
T Fusco ◽  
B Neichel
Keyword(s):  
Adaptive Optics ◽  
Blind Deconvolution ◽  
Simulated Data ◽  
Processing Technique ◽  
Post Processing ◽  
Very Large Telescope ◽  
Deconvolution Technique ◽  
Spread Function ◽  
Isolated Point ◽  
Correct Image

ABSTRACT One of the major limitations of using adaptive optics (AO) to correct image post-processing is the lack of knowledge about the system’s point spread function (PSF). The PSF is not always available as direct imaging on isolated point-like objects, such as stars. The use of AO telemetry to predict the PSF also suffers from serious limitations and requires complex and yet not fully operational algorithms. A very attractive solution is to estimate the PSF directly from the scientific images themselves, using blind or myopic post-processing approaches. We demonstrate that such approaches suffer from severe limitations when a joint restitution of object and PSF parameters is performed. As an alternative, here we propose a marginalized PSF identification that overcomes this limitation. In this case, the PSF is used for image post-processing. Here we focus on deconvolution, a post-processing technique to restore the object, given the image and the PSF. We show that the PSF estimated by marginalization provides good-quality deconvolution. The full process of marginalized PSF estimation and deconvolution constitutes a successful blind deconvolution technique. It is tested on simulated data to measure its performance. It is also tested on experimental AO images of the asteroid 4-Vesta taken by the Spectro-Polarimetric High-contrast Exoplanet Research (SPHERE)/Zurich Imaging Polarimeter (Zimpol) on the Very Large Telescope to demonstrate application to on-sky data.

Consortium for Adaptive Optics and Image Post-Processing

2008 ◽  
Author(s):  
Joseph Janni ◽  
Stuart Jefferies
Keyword(s):  
Adaptive Optics ◽  
Post Processing

scholarly journals Marginal blind deconvolution of adaptive optics retinal images

2011 ◽  
Vol 19 (23) ◽  
pp. 23227 ◽  
Author(s):  
L. Blanco ◽  
L. M. Mugnier
Keyword(s):  
Adaptive Optics ◽  
Blind Deconvolution ◽  
Retinal Images

Quality assessment for adaptive optics image post-processing by LoG domain matching

2018 ◽  
Vol 47 (11) ◽  
pp. 1111005
Author(s):  
牛 威 Niu Wei ◽  
郭世平 Guo Shiping ◽  
史江林 Shi Jianglin ◽  
邹建华 Zou Jianhua ◽  
张荣之 Zhang Rongzhi
Keyword(s):  
Quality Assessment ◽  
Adaptive Optics ◽  
Post Processing ◽  
Log Domain

Imaging through turbid media: post processing using blind deconvolution

2001 ◽  
Author(s):  
Stuart M. Jefferies ◽  
Kathy J. Schulze ◽  
Charles L. Matson ◽  
E. Keith Hege ◽  
Kurt Stoltenberg
Keyword(s):  
Blind Deconvolution ◽  
Turbid Media ◽  
Post Processing ◽  
Imaging Through Turbid Media

scholarly journals Wind-driven halo in high-contrast images

2020 ◽  
Vol 638 ◽  
pp. A98
Author(s):  
F. Cantalloube ◽  
O. J. D. Farley ◽  
J. Milli ◽  
N. Bharmal ◽  
W. Brandner ◽  
...  
Keyword(s):  
Adaptive Optics ◽  
Circumstellar Disks ◽  
Spectral Variation ◽  
Post Processing ◽  
Point Spread ◽  
Very Large Telescope ◽  
Spread Function ◽  
Processing Techniques ◽  
Fundamental Limitation ◽  
Adaptive Optics System

Context. The wind-driven halo is a feature that is observed in images that were delivered by the latest generation of ground-based instruments that are equipped with an extreme adaptive optics system and a coronagraphic device, such as SPHERE at the Very Large Telescope (VLT). This signature appears when the atmospheric turbulence conditions vary faster than the adaptive optics loop can correct for. The wind-driven halo is observed as a radial extension of the point spread function along a distinct direction (this is sometimes referred to as the butterfly pattern). When this is present, it significantly limits the contrast capabilities of the instrument and prevents the extraction of signals at close separation or extended signals such as circumstellar disks. This limitation is consequential because it contaminates the data for a substantial fraction of the time: about 30% of the data produced by the VLT/SPHERE instrument are affected by the wind-driven halo. Aims. This paper reviews the causes of the wind-driven halo and presents a method for analyzing its contribution directly from the scientific images. Its effect on the raw contrast and on the final contrast after post-processing is demonstrated. Methods. We used simulations and on-sky SPHERE data to verify that the parameters extracted with our method can describe the wind-driven halo in the images. We studied the temporal, spatial, and spectral variation of these parameters to point out its deleterious effect on the final contrast. Results. The data-driven analysis we propose provides information to accurately describe the wind-driven halo contribution in the images. This analysis confirms that this is a fundamental limitation of the finally reached contrast performance. Conclusions. With the established procedure, we will analyze a large sample of data delivered by SPHERE in order to propose post-processing techniques that are tailored to removing the wind-driven halo.

scholarly journals 9.5. The “double nucleus” of M31 in J, H, and K

1998 ◽  
Vol 184 ◽  
pp. 391-392
Author(s):  
P. Hinz ◽  
K. Hege ◽  
D. McCarthy ◽  
M. Lloyd-Hart ◽  
F. Melia
Keyword(s):  
Adaptive Optics ◽  
Near Infrared ◽  
Blind Deconvolution ◽  
Hubble Space Telescope ◽  
Wiener Filter ◽  
Diffraction Limit ◽  
Infrared Observations ◽  
Double Nucleus ◽  
Filter Algorithms ◽  
Adaptive Optics System

Hubble Space Telescope images of the nucleus of M31 show a double-peaked structure with the primary peak being offset from the center by approximately 0.5″. We observed the central 13″ of M31 in the J, H, and Ks passbands to determine the nuclear structure in the near-infrared. Observations were taken at the MMT Observatory, using a low-order adaptive optics system, FASTTRAC II (Gray et. al. 1995). The diffraction limit for the system is 0.25″ in K band. PSF images showed correction to 0.5″ FWHM. Uncorrected images showed the seeing to be about 1″. The images were deconvolved using several methods to check for consistency. We used Iterative-Blind Deconvolution, Richardson-Lucy, and Wiener filter algorithms, getting similar results for each. Measurements suggest the PSF in the deconvolved images is approximately 0.35″ FWHM.

Sign in / Sign up

Export Citation Format

Share Document