Comparative study of multi-look processing for phase map de-noising in digital Fresnel holographic interferometry

Area of Science:

• Optics and Photonics
• Image Processing
• Metrology

Background:

• Digital holographic interferometry (DHI) is crucial for precise measurements.
• Phase map de-noising is essential for accurate DHI analysis.
• Multi-look processing offers potential for improved phase map quality.

Purpose of the Study:

• To comparatively evaluate multi-look processing approaches for de-noising DHI phase maps.
• To assess the performance of various de-noising algorithms under simulated multi-look conditions.
• To identify optimal strategies for enhancing phase map accuracy in DHI.

Main Methods:

• Generation of 160 simulated phase fringe patterns with diverse fringe characteristics.
• Creation of 20 realistic noise realizations per pattern for multi-look simulation.
• Processing with 22 selected de-noising algorithms and evaluation of three multi-look strategies.
• Quantitative assessment using two distinct performance metrics.

Main Results:

• Consistent algorithm rankings were achieved across both evaluation metrics.
• A multi-look approach involving averaging before noise processing demonstrated superior performance.
• Two-dimensional windowed Fourier transform filtering emerged as the best-performing method.
• The block-matching 3D (BM3D) algorithm secured the second-highest performance ranking.

Conclusions:

• Averaging phase maps prior to de-noising is a highly effective multi-look strategy for DHI.
• 2D windowed Fourier transform filtering provides robust phase map de-noising in DHI applications.
• BM3D is a strong alternative for phase map de-noising, offering near-optimal results.