General construction of transform-domain filters, filtering methods for electronic speckle pattern interferometry,...

Area of Science:

• Optical Metrology
• Image Processing
• Signal Processing

Background:

• Speckle noise significantly degrades the quality of phase patterns in Electronic Speckle Pattern Interferometry (ESPI).
• Effective noise reduction is crucial for accurate quantitative analysis of ESPI measurements.
• Existing transform-domain filters offer partial solutions but can be further optimized.

Purpose of the Study:

• To present a general methodology for constructing transform-domain filters for speckle noise removal in ESPI.
• To review and evaluate established filters like Fourier Transform (FT), Wavelet Transform, Windowed Fourier Filtering, and Localized FT filters.
• To propose and validate three novel transform-domain filters designed for enhanced speckle noise reduction in ESPI.

Main Methods:

• Development of a general framework for designing transform-domain filters.
• Comparative performance analysis of existing filters using various test cases and Mean Square Error (MSE).
• Design and implementation of three new filters by optimizing input image and filter functions.

Main Results:

• Established filters were reviewed and their performance evaluated.
• Three new transform-domain filters were developed based on the general method.
• The proposed filters demonstrated superior speckle noise removal capabilities compared to existing methods in ESPI phase patterns.

Conclusions:

• The general method provides a flexible approach to designing effective speckle noise filters for ESPI.
• The newly proposed filters offer significant improvements in noise reduction, enhancing the reliability of ESPI measurements.
• This work contributes advanced image processing techniques for optical metrology applications.