Gradient descent algorithm applied to wavefront retrieval from through-focus images by an extreme ultraviolet...

Area of Science:

• Optical imaging
• Wavefront sensing
• Computational optics

Background:

• Gradient descent algorithms are standard for wavefront retrieval in coherent/incoherent imaging.
• Accurate wavefront retrieval requires modeling partial coherence and object transmittance.
• Simulating partially coherent imaging is computationally intensive, slowing optimization.

Purpose of the Study:

• To extend gradient descent wavefront retrieval for partially coherent illumination.
• To accelerate computation in wavefront retrieval for complex imaging systems.
• To demonstrate the algorithm's efficacy in extreme ultraviolet microscopy.

Main Methods:

• Incorporated partial coherence and object transmittance into gradient descent.
• Utilized Fourier transform and eigenfunction decomposition for computational acceleration.
• Applied the extended algorithm to retrieve wavefronts from through-focus images.

Main Results:

• Successfully retrieved a field-dependent wavefront using the extended algorithm.
• Demonstrated qualitative agreement between retrieved wavefront and lens design.
• Achieved accelerated computation through eigenfunction decomposition.

Conclusions:

• The extended gradient descent algorithm effectively retrieves wavefronts under partially coherent illumination.
• Eigenfunction decomposition significantly reduces computation time for complex imaging simulations.
• The method is suitable for wavefront characterization in extreme ultraviolet microscopy.