2-D Deformation Measurement by Digital
Image Correlation (DIC)
Contact: Jean-José Orteu (Jean-Jose.Orteu@enstimac.fr) – Laurent Robert (Laurent.Robert@enstimac.fr)
Digital Image Correlation (DIC)
time t0 time t1 time t2
Purpose: Deformation measurement from a The gray levels of image patches are
sequence of images taken by a single camera matched, assuming that they present
some similarity, by using Digital Image
• Full-field measurement Correlation (DIC).
The image correlation technique is
• Planar specimen only
efficient on textured objects. If the
• No out-of-plane motion objects are not enough textured
naturally, it only takes a few seconds to
mark the surface of the object with a
random pattern of spray paint.
The similarity of the image patches is computed using a correlation criterion (similarity score).
The specimen is deformed: the correlation algorithm must also deform the correlation window
(adaptive correlation window shape). The displacement transformation is modeled (parametric
representation) and the parameters of the transformation are estimated using optimization algorithms
Time t Time t’
To achieve sub-pixel accuracy the correlation algorithms use gray value interpolation:
• Optimization algorithms require the criterion to be continuous over its parameter space
• Images are discrete: we need to reconstruct the continuous information by means of
• B-Splines are a good choice for that purpose
Raw image B-spline interp.
Bi-linear B-spline interp. B-spline interp.
Bi-cubic B-spline interp.
“Punch” deformation The DIC technique gives the displacement field. The strain field
is computed by numerical derivation of the displacement field.
Initial image Deformed image
y Ux Uy
image 10 degrees
Strain field (computed by numerical derivation)
ε xx ε
yy xy ε
Results obtained using Vic-2D® software Results obtained using Vic-2D® software
Tensile test on an elastomer sample Tensile test on a polymer (PEI) sample (results obtained using Vic-2D® software)
Strain field εxx εyy εxy