Description and simulation of an active imaging technique utilizing two speckle fields: iterative reconstructors.

Area of Science:

• Optics
• Image Reconstruction
• Computational Imaging

Background:

• Laser speckle patterns contain object shape information.
• Previous root-matching techniques for image reconstruction are slow and noise-sensitive.
• Advanced algorithms are needed for efficient and accurate speckle-based imaging.

Purpose of the Study:

• To investigate alternative algorithms for laser speckle imaging.
• To improve the speed and noise robustness of object shape reconstruction.
• To achieve high-quality image reconstruction from speckle data.

Main Methods:

• Applied phase retrieval, expectation maximization, and statistical maximization algorithms.
• Utilized two-field data (speckle intensity patterns and interference) at multiple wavelengths.
• Evaluated algorithm performance for reconstructing complex objects.

Main Results:

• Phase retrieval and expectation maximization were most effective for complex objects (>10 pixels).
• High-quality images were formed using three sets of two-wavelength data.
• Expectation maximization showed good results with single-wavelength data (≥3 frames).
• Phase retrieval produced good images using only object autocorrelation.

Conclusions:

• Phase retrieval and expectation maximization significantly advance laser speckle imaging.
• These methods provide robust and high-quality object shape reconstruction.
• The study demonstrates improved imaging capabilities for complex objects.