High dynamic range coherent imaging using compressed sensing.

Area of Science:

• Optics and Photonics
• Digital Imaging
• Signal Processing

Background:

• Lensless Fourier transform holography (FTH) and coherent diffraction imaging (CDI) often use beamstops, which block high intensities but cause loss of low-frequency information.
• This data loss hinders high-quality image reconstruction in both FTH and CDI.

Purpose of the Study:

• To demonstrate image recovery from high-frequencies only, overcoming the beamstop problem in FTH and CDI.
• To apply compressed sensing (CS) techniques for reconstructing images from incomplete holographic and diffraction data.

Main Methods:

• Utilized compressed sensing (CS) reconstruction for captured holograms in FTH, applicable to various FTH configurations.
• Developed a novel phase retrieval procedure for CDI, integrating Fienup's hybrid input-output (HIO) method with CS.

Main Results:

• Successfully recovered images from high-frequencies alone, effectively addressing the beamstop-induced data loss.
• Demonstrated the robustness and effectiveness of CS-based reconstruction through numerical simulations and experimental validation for both FTH and CDI.

Conclusions:

• Image reconstruction from high-frequencies is feasible for sparse objects, even with missing low-frequency information caused by beamstops.
• The proposed CS-based methods offer a robust solution for improving image quality in FTH and CDI by handling incomplete measurements.