Efficient and automated initial value estimation in digital image correlation for large displacement, rotation, and...

Area of Science:

• * Digital Image Correlation (DIC)
• * Optical Measurement Techniques
• * Signal Processing

Background:

• * Initial value estimation is crucial for digital image correlation (DIC) calculations.
• * Fourier-Mellin transform-based cross correlation (FMT-CC) offers rotation and scale invariance but fails with large displacements.
• * Existing FMT-CC methods lack robustness for significant image transformations.

Purpose of the Study:

• * To develop an automated and efficient initial value estimation method for DIC.
• * To address the limitations of FMT-CC with large displacements, rotations, and isotropic scaling.
• * To enhance the robustness and computational efficiency of DIC initial value estimation.

Main Methods:

• * Investigated the relationship between subset size and maximal displacement in FMT-CC.
• * Proposed a strategy for setting subset size based on estimated displacement.
• * Introduced a multi-scale search method to improve efficiency for large displacements.

Main Results:

• * The proposed method enables rapid and automated initial value estimation under large displacement, rotation, and scaling.
• * The strategy of adjusting subset size enhances the robustness of FMT-CC.
• * The multi-scale search method achieves a computational efficiency approximately one order of magnitude higher than traditional FMT-CC.

Conclusions:

• * The developed method significantly improves the performance of initial value estimation in DIC.
• * The approach overcomes the limitations of traditional FMT-CC for complex image transformations.
• * This advancement facilitates more accurate and efficient DIC analysis in challenging scenarios.