Randomized apertures: high resolution imaging in far field.

Area of Science:

• Optical astronomy
• Image reconstruction
• Computational imaging

Background:

• Extremely large telescope (ELT) designs face challenges with optical aberrations.
• Randomly varying subapertures offer a potential solution for aberration correction.
• Super-resolution imaging requires advanced deconvolution techniques.

Purpose of the Study:

• To explore the feasibility of an ELT design using ill-figured, randomly varying subapertures.
• To demonstrate super-resolution imaging capabilities beyond the diffraction limit.
• To develop and validate a robust image restoration framework for such systems.

Main Methods:

• Laboratory-scaled system demonstrating white light binary point source reconstruction.
• Application of a restoration framework combining support vector machine (SVM) based lucky imaging.
• Integration of non-negative matrix factorization (NMF) based multiframe blind deconvolution.
• Modeling of the experimental system for sub-diffraction-limited performance analysis.

Main Results:

• Successful reconstruction of a binary point source separated by 2.5 times the diffraction limit.
• Demonstration of the system's ability to handle an unknown, varying point spread function.
• Validation of the approach through simulation of sub-diffraction-limited performance with multiple point sources.

Conclusions:

• The proposed ELT design with ill-figured, randomly varying subapertures is viable for super-resolution imaging.
• The combined SVM lucky imaging and NMF multiframe blind deconvolution framework effectively restores images from speckle data.
• This approach holds promise for achieving unprecedented resolution in future astronomical instruments.