High-fidelity pixel-super-resolved complex field reconstruction via adaptive smoothing.

Area of Science:

• Optics and Photonics
• Digital Imaging
• Computational Imaging

Background:

• Lensless digital holographic imaging faces resolution limits due to sampling.
• Pixel super-resolution (PSR) aims to overcome these limits but suffers from non-convexity and noise sensitivity.
• Existing PSR algorithms can converge to suboptimal solutions and are prone to noise interference.

Purpose of the Study:

• To develop a robust pixel super-resolution (PSR) algorithm for high-fidelity reconstruction in lensless digital holographic imaging.
• To address the challenges of non-convexity and noise in PSR phase retrieval.
• To enhance the resolution of holographic images towards the diffraction limit.

Main Methods:

• Proposed a heuristic PSR algorithm with adaptive smoothing (AS-PSR).
• Implemented automatic adjustment of intensity constraints on the estimated field.
• Verified the method experimentally within a coherent modulation phase retrieval framework.

Main Results:

• AS-PSR achieved high-fidelity image reconstruction.
• The algorithm effectively located the optimal solution and converged with high quality.
• A twofold improvement in resolution was experimentally demonstrated.
• The method showed robustness against noise inherent in the up-sampling procedure.

Conclusions:

• The adaptive smoothing pixel super-resolution (AS-PSR) algorithm enhances reconstruction quality and pushes resolution limits in lensless holographic imaging.
• AS-PSR offers a robust solution to the non-convexity and noise issues in PSR.
• The algorithm is applicable to other phase retrieval methods utilizing alternating projections.