“…There is always a compromise between the standard displacement resolution and the spatial resolution (e.g., the size of the interrogation volume), namely, the larger the interrogation volume, the smaller the displacement resolution. This result is very general to many optical techniques utilized to measure kinematic elds, in particular digital image correlation [17,18] and digital volume correlation [19].…”
Most of the norms used in the eld of digital image (and volume) correlation to register two images (or volumes) lead to ill-posed problems. One of the frequent solutions is to enforce a restricted kinematics requiring a compromise between the richness of the solution (i.e., the spatial resolution) and the measurement uncertainty.An alternative route is to use a displacement norm that permits to alleviate this compromise by the means of a mechanical regularization used when the gray levels do not give enough information. It is then possible to compute a displacement vector for each pixel or voxel, inducing lower residuals (in terms of experimental data) while decreasing the noise sensitivity.The resolution performance of these di erent approaches is discussed, and compared for the analysis of a tensile test on a cast iron specimen based on a pair of tomographic images. As representative reconstructed volumes lead to a large number of degrees of freedom, a dedicated GPU computational strategy has been developed and implemented.Keywords: Global digital image correlation; GPU; Regularization.
“…There is always a compromise between the standard displacement resolution and the spatial resolution (e.g., the size of the interrogation volume), namely, the larger the interrogation volume, the smaller the displacement resolution. This result is very general to many optical techniques utilized to measure kinematic elds, in particular digital image correlation [17,18] and digital volume correlation [19].…”
Most of the norms used in the eld of digital image (and volume) correlation to register two images (or volumes) lead to ill-posed problems. One of the frequent solutions is to enforce a restricted kinematics requiring a compromise between the richness of the solution (i.e., the spatial resolution) and the measurement uncertainty.An alternative route is to use a displacement norm that permits to alleviate this compromise by the means of a mechanical regularization used when the gray levels do not give enough information. It is then possible to compute a displacement vector for each pixel or voxel, inducing lower residuals (in terms of experimental data) while decreasing the noise sensitivity.The resolution performance of these di erent approaches is discussed, and compared for the analysis of a tensile test on a cast iron specimen based on a pair of tomographic images. As representative reconstructed volumes lead to a large number of degrees of freedom, a dedicated GPU computational strategy has been developed and implemented.Keywords: Global digital image correlation; GPU; Regularization.
“…[44], ring artifacts degrade the accuracy of DVC, while beam hardening and its correction had negligible effects. However, more investigation is warranted to assess the impact of the ring artifact reduction on accuracy.…”
Section: Sample Preparationmentioning
confidence: 97%
“…Instead of defining a regular Cartesian grid, they segment the image to identify individual sand grains and center a subset on each sand grain. They used six translation and rotational DOF, assuming the grains were rigid, and computed DVC for 50,000 grains at 1 second per 27 3 Instead of the subset based DVC presented here, Roux et al introduced a finiteelement based DVC [37,44,67]. This approach defines a finite-element mesh on the image and computes the deformation of the entire mesh at once, instead of computing each subset independently.…”
mentioning
confidence: 99%
“…They studied small 100 3 to 288 3 voxel regions using elements from 6 3 to 18 3 voxels. Using this finite-element DVC, Limodin et al [44] studied the effects of CT scanner artifacts and the spurious strain induced by thermal expansion caused by the X-rays in the scanner. Recently, Leclerc et al [41] presented a voxel-scale DVC where the displacement of each voxel is determined (i.e., a subset of size 1) with the addition of the regularization that local behavior is elastic, to make the problem well-posed.…”
My colleagues in scientific computing have been very helpful in discussing ideas about optimization, image processing, GPU computing, Python, and numerous other topics. Particularly, I am appreciative of conversations with Russ Hewett,
“…Tomography images are reconstructed from a set of radiographs, a procedure producing artifacts / noise that must be taken into account when digital volume correlation (DVC) measurements are to be performed [9].…”
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.