Transport of intensity equation-based computational phase retrieval for robust quantitative phase imaging

Area of Science:

• Computational imaging
• Optical physics
• Image reconstruction

Background:

• Quantitative phase imaging (QPI) methods like the transport of intensity equation (TIE) are valuable for non-interferometric imaging.
• Standard TIE methods face limitations including sensitivity to uniform intensity, boundary conditions, and blurring at large defocus distances.

Purpose of the Study:

• To develop an iterative phase recovery algorithm that overcomes the limitations of traditional TIE.
• To enhance the accuracy and robustness of QPI, particularly for challenging experimental conditions.

Main Methods:

• An iterative phase recovery algorithm was developed, integrating a TIE phase-compensation solution.
• The algorithm minimizes the difference between experimental and theoretical axial intensity derivatives using angular spectrum propagation.
• Physics-based priors derived from TIE were incorporated into the iterative reconstruction.

Main Results:

• Simulations demonstrated the algorithm's validity across various non-uniform intensity, noise, boundary condition, and defocus distance scenarios.
• Experimental verification confirmed the method's effectiveness for large defocus distances.
• The proposed method showed robustness when compared to standard TIE solutions.

Conclusions:

• The proposed iterative algorithm offers a fast and accurate phase recovery solution for QPI.
• This method effectively addresses key limitations of the transport of intensity equation, enhancing its practical applicability.