Phase retrieval and Zernike decomposition using measured intensity data and the estimated electric field

Area of Science:

• Optical Engineering
• Computational Optics
• Image Processing

Background:

• Accurate optical system characterization is crucial for high-performance imaging.
• Traditional phase retrieval methods often require image defocusing or suffer from phase unwrapping limitations.
• Zernike coefficients are a standard representation for optical aberrations.

Purpose of the Study:

• To introduce a new direct search phase retrieval technique for determining optical prescriptions using Zernike coefficients.
• To leverage electric field (E-field) estimates alongside intensity data for improved phase retrieval.
• To overcome limitations associated with traditional phase diversity and phase unwrapping methods.

Main Methods:

• Utilizes a direct search algorithm combined with Gerchberg-Saxton (GS) algorithm for iterative E-field estimation.
• Employs E-field patterns, which are less correlated than intensity patterns, for enhanced Zernike coefficient estimation.
• Correlates estimated E-field with modeled E-field to derive Zernike coefficients, minimizing error between measured and modeled intensity data.

Main Results:

• The proposed technique provides Zernike coefficient estimates without requiring defocused images.
• Numerical analysis indicates E-field patterns offer more information for Zernike coefficient estimation than intensity patterns alone.
• The algorithm successfully avoids local minima in simulations across a range of aberrations.
• Experimental validation confirms the effectiveness of the phase retrieval technique.

Conclusions:

• The novel phase retrieval method offers a robust and effective approach for determining optical prescriptions.
• Using E-field estimates circumvents phase unwrapping issues inherent in some GS algorithm applications.
• This technique enhances the accuracy and reliability of Zernike coefficient estimation for optical systems.