Comparative analysis of image pre-filtering techniques for phase-shifted noise-affected interferograms

Area of Science:

• Optical Metrology
• Interferometry
• Signal Processing

Background:

• Phase-shifting techniques are crucial for extracting quantitative phase information from interferograms.
• Noise in interferograms significantly degrades measurement accuracy and parameter retrieval.
• Traditional methods lack robust noise reduction strategies.

Purpose of the Study:

• To comparatively analyze Fourier, windowed Fourier, and wavelet filtering for noise reduction in phase-shifting interferometry.
• To evaluate the effectiveness of these filtering techniques in extracting phase information and measuring parameters.
• To assess the impact of noise and second-order harmonics on measurement accuracy.

Main Methods:

• Simulating interferograms with additive and multiplicative noise.
• Applying Fourier, windowed Fourier, and wavelet filtering as pre-filtering strategies.
• Analyzing the impact of second-order harmonics in the presence of noise.
• Experimental validation using coherent (Talbot) and incoherent (Lau) grating shearing interferometers.

Main Results:

• Wavelet filtering demonstrated superior performance in noise reduction and phase information extraction compared to Fourier and windowed Fourier methods.
• Accurate measurement of defocusing error was achieved using the investigated filtering strategies.
• The analysis confirmed the effectiveness of pre-filtering techniques in enhancing measurement precision.

Conclusions:

• Pre-filtering, particularly wavelet filtering, is essential for improving the accuracy of phase-shifting interferometry.
• These techniques offer robust solutions for noise mitigation in optical metrology applications.
• The study provides valuable insights for selecting appropriate filtering methods for interferogram analysis.