Strain fields measurement using frequency domain Savitzky–Golay filters in digital image correlation

Abstract

Savitzky–Golay (SG) filter is an effective and convenient gradient calculation method employed for full-field strain measurement in digital image correlation (DIC). Currently, the strain field can be conveniently obtained by moving smoothing filtering with SG filters. This is a method of spatial filtering, which offers the advantages of easy implementation and high accuracy. This study proposed a strain calculation method based on frequency domain SG (FD-SG) filtering. Prior to performing FD-SG filtering, data extension operations involving outer-boundary padding and zero-padding were performed on the displacement field data. Similarly, the SG filters template was extended with zero-padding and circularly shift operations. Subsequently, FD-SG filters were generated by applying the fast Fourier transform (FFT) to the expanded SG template. Thereafter, FD-SG filtering was implemented through the multiplication of the displacement in the Fourier-domain by an FD-SG filter. Finally, the strain field was obtained via inverse FFT and valid data interception operation. The simulation and practical experiments confirmed the equivalent accuracy exhibited by FD-SG and spatial domain SG filtering. Thus, the proposed FD-SG filtering method has great potential for real-time strain measurement in DIC.