End-to-end aperture layout optimization and image restoration for optical sparse aperture systems

Area of Science:

• Astronomy and Astrophysics
• Optical Engineering
• Computational Imaging

Background:

• Optical sparse aperture (OSA) imaging offers improved spatial resolution with reduced size, weight, and cost for telescopes.
• Current OSA research often separates aperture layout design and image restoration, leading to design inefficiencies.
• Existing methods may not fully leverage the information captured by sparse apertures for optimal image reconstruction.

Purpose of the Study:

• To develop an end-to-end design framework for optical sparse aperture systems.
• To simultaneously optimize aperture layout and neural network parameters for image restoration.
• To achieve excellent imaging quality in OSA systems through integrated design.

Main Methods:

• Proposed an end-to-end design framework integrating OSA aperture layout and neural network image restoration.
• Optimized both aperture layout parameters and neural network parameters within a unified framework.
• Simulated a simplified OSA system on geostationary orbit using the developed framework.

Main Results:

• The integrated framework achieved excellent imaging quality.
• Mid-frequency image information captured by OSA systems proved more beneficial for network processing than sparse high-frequency information.
• A simplified six-sub-aperture OSA system demonstrated comparable performance to a 12m single-aperture system.

Conclusions:

• Simultaneous optimization of aperture layout and image restoration is effective for OSA systems.
• Mid-frequency information is crucial for effective image restoration in OSA systems.
• A simplified OSA system can achieve high-performance imaging, comparable to much larger single apertures.