Automatic digital processing in speckle photography: comparison of two algorithms

Area of Science:

• Optics and Photonics
• Digital Image Processing
• Metrology

Background:

• Speckle photography is a technique used for non-contact measurement of surface deformation and displacement.
• Accurate analysis of speckle photography fringes is crucial for reliable metrology.
• Existing algorithms vary in computational efficiency and precision.

Purpose of the Study:

• To compare the performance of two distinct algorithms for analyzing speckle photography fringes.
• To evaluate algorithm accuracy and computational time in fringe spacing and orientation determination.
• To assess algorithm robustness against data degradation caused by noise.

Main Methods:

• Development and implementation of two algorithms for speckle fringe analysis.
• Computer simulation of speckle photography data, incorporating controlled levels of noise.
• Quantitative assessment of fringe spacing and orientation accuracy for each algorithm.
• Measurement of computer run-time for each algorithm under simulated conditions.

Main Results:

• Algorithm A demonstrated higher accuracy in determining fringe spacing and orientation compared to Algorithm B.
• Algorithm B exhibited a faster computer run-time, making it more computationally efficient.
• Both algorithms showed a decrease in accuracy with increasing levels of data noise, with Algorithm A being more resilient.

Conclusions:

• Algorithm A is recommended for applications requiring high accuracy in speckle fringe analysis, especially with noisy data.
• Algorithm B is suitable for real-time applications where computational speed is prioritized over absolute accuracy.
• Further research could explore hybrid approaches to combine the strengths of both algorithms.