Single-shot fringe pattern phase retrieval using improved period-guided bidimensional empirical mode decomposition...

Area of Science:

• Optical Metrology
• Image Processing
• Signal Processing

Background:

• Fringe pattern analysis is crucial for optical measurement techniques like interferometry.
• Environmental and system noise in experimental fringe patterns degrades analysis accuracy.
• Pre-processing is essential for effective phase retrieval from interferograms.

Purpose of the Study:

• To develop a fully automatic and adaptive fringe pattern pre-processing technique.
• To enhance the performance and applicability of the Period Guided Bidimensional Empirical Mode Decomposition (PGBEMD) algorithm.
• To create a comprehensive and robust fringe pattern analysis pipeline.

Main Methods:

• Proposed an improved Period Guided Bidimensional Empirical Mode Decomposition (iPGBEMD) algorithm.
• Addressed low fringe amplitude image decomposition issues with three novel solutions.
• Integrated acceleration methods for computational efficiency.
• Combined iPGBEMD with the Hilbert Spiral Transform for a complete analysis path.

Main Results:

• Significantly increased the application range of PGBEMD.
• Reduced computation time several-fold compared to the original PGBEMD.
• Successfully handled fringe patterns with very low fringe amplitudes.
• Demonstrated effective fringe pattern analysis through numerical simulations and experimental data.

Conclusions:

• The iPGBEMD algorithm offers a robust and efficient solution for fringe pattern pre-processing.
• The integrated analysis path provides consistent and versatile fringe pattern analysis.
• The proposed method outperforms existing techniques in quality and effectiveness for optical measurements.