Automatic fringe pattern enhancement using truly adaptive period-guided bidimensional empirical mode decomposition

Area of Science:

• Optical Metrology
• Image Processing
• Signal Processing

Background:

• Fringe patterns are vital for optical full-field testing methods like interferometry.
• These patterns are often degraded by noise, uneven illumination, and low contrast.
• Pre-processing for fringe pattern enhancement is essential for accurate measurement results.

Purpose of the Study:

• To develop an automatic, robust, and effective fringe pattern enhancement method.
• To address noise, background illumination, and low contrast issues in fringe patterns.
• To improve the accuracy of optical measurement results by enhancing fringe patterns before phase map calculation.

Main Methods:

• A novel period-guided bidimensional empirical mode decomposition (PG-BEMD) algorithm is proposed.
• Fringe period spatial distribution is estimated using a windowed approach for adaptive decomposition.
• Block matching 3D filtering is employed for noise removal prior to decomposition.

Main Results:

• The PG-BEMD algorithm successfully extracts the fringe term in a single component, avoiding mode mixing.
• The method demonstrates robustness in noise removal and effective background term elimination.
• Validation using numerical simulations and experimental data confirms the approach's effectiveness and versatility.

Conclusions:

• The proposed PG-BEMD method offers a significant advancement in fringe pattern enhancement for optical measurements.
• The adaptive decomposition and integrated noise reduction provide a robust solution for degraded fringe patterns.
• This technique simplifies signal reconstruction and enhances the reliability of optical testing outcomes.