Adaptive step-size strategy for noise-robust Fourier ptychographic microscopy

Area of Science:

• Optics and Imaging
• Computational Science

Background:

• Ptychographic imaging techniques like PIE and ePIE offer flexibility and efficiency.
• Image noise significantly degrades the stability and reconstruction quality of these methods.
• The issue is often linked to phase retrieval's non-convexity, but step-size choice is critical.

Purpose of the Study:

• To develop a noise-robust adaptive step-size strategy for incremental gradient methods in ptychographic imaging.
• To enhance the stability and reconstruction quality of Fourier ptychographic microscopy under noisy conditions.

Main Methods:

• Introduced an adaptive step-size strategy that dynamically reduces step size when progress stagnates.
• Applied and validated the strategy using synthetic and real data in Fourier ptychographic microscopy.

Main Results:

• The adaptive step-size strategy significantly enhances reconstruction stability and robustness against noise.
• The approach maintains the rapid initial convergence of existing methods like PIE and ePIE.
• Demonstrated improved performance without requiring prior knowledge of noise statistics.

Conclusions:

• The proposed adaptive step-size strategy offers a simple, nonparametric, and effective solution for robust ptychographic imaging.
• This technique is highly promising for improving Fourier ptychographic microscopy in noisy environments.
• The method balances performance gains with low computational complexity.