A Study on an Analysis Speed Enhancement Algorithm for the Digital Image Correlation Method Based on Fast Fourier Transform

Digital Image Correlation (DIC) is a measurement method based on image analysis that calculates the correlation between the original reference image and the deformed one. However, an extensive iterated calculation process is required to trace every point on the deformed specimen in the area of interest. The Fast Fourier Transform (FFT) has been adopted to reduce the calculation time, but the FFT method easily loses tracking for large deformation. A new FFT-based correlation analysis algorithm with displacement feedback is proposed to improve the calculation speed of the DIC system and to prevent the correlation coefficient from declining. The calculation rate using the conventional DIC method was compared with that using our new feedback FFT-based approach. This paper demonstrates that the calculation speed of the proposed method was about three times faster than that of the conventional technique. In addition, no tracking failure comes out by our new algorithm even though only FFT-DIC without feedback updating lost tracking over 10% strain.