Convolutional Neural Networks Approach for Solar Reconstruction in SCAO Configurations.

Area of Science:

• Astronomy and Astrophysics
• Optical Engineering
• Computer Science

Background:

• Atmospheric turbulence distorts light from astronomical objects, necessitating adaptive optics (AO) for clear Earth-based telescope observations.
• Traditional AO correction relies on reconstruction algorithms, with neural networks showing promise.

Purpose of the Study:

• To develop and evaluate a novel convolutional neural network approach for correcting image aberrations in solar observations using AO.
• To address the unique challenges of solar AO, which differ from nocturnal astronomical observations.

Main Methods:

• A convolutional neural network approach was employed for adaptive optics correction.
• A specific reconstruction algorithm, "Shack-Hartmann reconstruction with deep learning on solar-prototype" (proto-HELIOS), was developed and tested.
• The method was applied to fixed solar images in Single-Conjugate Adaptive Optics (SCAO) configurations.

Main Results:

• The proto-HELIOS algorithm achieved an average precision of 85.39% in reconstructing corrected solar images.
• The convolutional approach demonstrated effectiveness in addressing solar image aberrations.

Conclusions:

• The developed deep learning technique shows significant potential for correcting aberrations in solar adaptive optics.
• Further research is encouraged to refine these techniques for comprehensive solar region coverage.