Subpixel displacement and deformation gradient measurement using digital image/speckle correlation (DISC)

Abstract

Experimental solid mechanics relies heavily on surface displacement and deformation gradient measurements. Digital image/ speckle correlation (DISC) uses digital image processing to resolve displacement and deformation gradient fields. The practical implementation of DISC involves important challenges such as computation complexity and the discrepancy of the sensitivities and accuracies claimed in previous studies. We develop an iterative, spatial-gradient based algorithm, which uses only first-order spatial derivatives of the images before and after deformation. Simulated images are then used to verify this algorithm, as well as to study the impact of speckle size on the accuracy. Based on these simulations, the sensitivity of DISC to displacement and deformation gradient, as well as an optimal speckle size for optimal accuracy, is recommended. The algorithm is then calibrated using rigid body translation and rotation, and an application of DISC to thermomechanical diagnostics of electronic packaging is also presented. © 2001 Society of Photo-Optical Instrumentation Engineers.

Precise In-Process Strain Measurements for the Investigation of Surface Modification Mechanisms

“…The simulations revealed a speckle size of 2-5 pixels and an evaluation window size of more than 6 pixels as the optimum for the evaluation of the speckle pattern, if the two-dimensional Gaussian interpolation method is used for the DISC calculation. These values are similar to the results presented by Zhou and Goodson for a fixed number of speckles within the evaluation window [12]. For this reason, a speckle size of 3 pixels and an evaluation window size of 10 pixels are applied in the next considerations as well as in the measurement system.…”

“…In order to obtain the global deformation field, the evaluation window scans across the whole field of view, whereby the local deformation Shiftx,y is stored for each scan position (px, py). Due to the almost Gaussian profile of the cross-correlation function, a twodimensional Gaussian curve is fitted to the correlation data to determine the maximum position with subpixel resolution [12]. After a corresponding low-pass filtering, the strain and thus the stress can be calculated from the deformation field by gradient formation.…”

Micropatterned mammalian cells exhibit phenotype-specific left-right asymmetry

“…A fast Fourier transform (FFT)-based method was first utilized to match regions of two sequential phase contrast images. The calculated displacement values were then used as inputs for a more accurate estimation of displacement fields at a subpixel level, using a second-order image correlation algorithm described previously (26). To evaluate the bias of cell migration, the obtained velocity field was further projected in the circumferential (V θ ) and radial (V r ) direction.…”

“…These data suggest that significant inhibition of the random walk of the cells through physical cell-cell contact was necessary for the cells to express subtle directional migration and biased alignment. The cells were labeled and tracked along the inner and outer ring pattern (Movie S4), and the cell migration velocity was estimated by digital image correlation (26). The speed of migration was higher at the inner and outer ring boundaries than within the interior region (25 μm∕hr vs. 10 μm∕hr; p < 0.05).…”

Measurement Precision and Spatial Resolution with Kriging Digital Image Correlation

“…For example, some algorithms [14] to [16] are employed to generate synthetic images that mimic real patterns of displacement in both Fourier [17] and spatial domains [18]. Artificial displacements may be applied to a piece of experimental speckle image [19] for the same purpose.…”