Multiframe blind deconvolution of heavily blurred astronomical images

Area of Science:

• Astronomy and Astrophysics
• Image Processing
• Computational Science

Background:

• Image restoration is crucial for analyzing astronomical data.
• Blind deconvolution algorithms are needed when the point-spread function (PSF) is unknown.
• Existing methods often require prior knowledge or specific constraints.

Purpose of the Study:

• To propose a novel multichannel blind deconvolution algorithm for image restoration.
• To develop an algorithm robust to various distortions and without restrictive assumptions.
• To validate the algorithm's performance using simulations and real astronomical data.

Main Methods:

• Incorporation of maximum-likelihood image restoration with multiple PSF estimates into the Ayers-Dainty algorithm.
• Development of an iterative approach that assumes only positivity of images and PSFs.
• No reliance on cost functions, input parameters, or prior probability distributions.

Main Results:

• Demonstrated convergence of the algorithm to positive estimates of the object and PSFs.
• Successful processing of real telescope data from multiple satellites, showcasing practical applicability.
• Analysis of convergence dependence on the number of processed frames in multiframe scenarios.

Conclusions:

• The proposed multichannel blind deconvolution algorithm offers a robust and parameter-free approach to image restoration.
• The algorithm's convergence is proven, particularly in multiframe cases, enhancing reliability for astronomical imaging.
• Novel metrics for evaluating performance and a stopping rule are introduced, optimizing the iterative restoration process.