Phase-gradient-based initialization for lensless quantitative phase imaging

Area of Science:

• Optical Imaging
• Computational Imaging
• Phase Retrieval

Background:

• Quantitative phase imaging faces challenges with transparent samples exhibiting large phase variations.
• Phase retrieval from intensity-only measurements is a non-convex problem, often leading to non-physical solutions with conventional algorithms.

Purpose of the Study:

• To develop a robust initialization strategy for coded ptychographic imaging that overcomes limitations in large-phase regimes.
• To enable stable and physically correct phase reconstruction for transparent samples with significant phase variations.

Main Methods:

• A phase-gradient-based initialization strategy is proposed for coded ptychographic imaging.
• Local phase gradients are extracted from intensity patterns as lateral displacements.
• A globally consistent initial phase is constructed using displacement-derived information.

Main Results:

• The proposed initialization strategy enables stable and physically correct phase reconstruction.
• Successful reconstruction is demonstrated in large-phase regimes where random initialization fails.
• The method achieves accurate results without requiring increased measurement redundancy.

Conclusions:

• Phase-gradient-based initialization provides essential first-order physical constraints for phase retrieval.
• This strategy significantly improves the reliability of quantitative phase imaging for challenging samples.
• The approach offers a breakthrough for analyzing transparent materials with complex optical properties.