Anisoplanatic deconvolution of adaptive optics images

Area of Science:

• Astronomy and Astrophysics
• Image Processing

Background:

• Adaptive optics (AO) systems in astronomy aim to correct atmospheric turbulence.
• Atmospheric turbulence causes anisoplanatism, a spatially varying point-spread function (PSF), degrading AO image quality.
• Maximum-likelihood deconvolution is a powerful image restoration technique but struggles with spatially variant PSFs.

Purpose of the Study:

• To develop a modified maximum-likelihood deconvolution method for astronomical adaptive optics images.
• To address the challenge of anisoplanatism by parametrizing the spatially variant PSF.
• To enable simultaneous optimization of the PSF and the deconvolved image.

Main Methods:

• A modified maximum-likelihood deconvolution algorithm is presented.
• The method parametrizes the anisoplanatic character of the point-spread function (PSF).
• Simultaneous optimization of the spatially variant PSF and the deconvolved image is performed.

Main Results:

• The algorithm demonstrates perfect cancellation of anisoplanatism effects down to numerical precision in ideal conditions.
• Performance is quantified in the presence of Poissonian noise for both crowded and noncrowded stellar fields.
• Two deconvolution modes were explored: using pixel values or stellar field parameters.

Conclusions:

• The developed method offers a robust solution for deconvolving astronomical adaptive optics images affected by anisoplanatism.
• The simultaneous optimization approach significantly enhances image quality.
• The algorithm's performance is well-characterized across different stellar field densities and noise levels.