When Deep Learning Meets Digital Image Correlation - Archive ouverte HAL Access content directly
Journal Articles Optics and Lasers in Engineering Year : 2021

When Deep Learning Meets Digital Image Correlation

Kamel Abdelouahab
  • Function : Author
François Berry
Benoît Blaysat
Michel Grediac
  • Function : Author
  • PersonId : 1038622


Convolutional Neural Networks (CNNs) constitute a class of Deep Learning models which have been used in the recent past to resolve many problems in computer vision, in particular optical flow estimation. Measuring displacement and strain fields can be regarded as a particular case of this problem. However, it seems that CNNs have never been used so far to perform such measurements. This work is aimed at implementing a CNN able to retrieve displacement and strain fields from pairs of reference and deformed images of a flat speckled surface, as Digital Image Correlation (DIC) does. This paper explains how a CNN called StrainNet can be developed to reach this goal, and how specific ground truth datasets are elaborated to train this CNN. The main result is that StrainNet successfully performs such measurements, and that it achieves competing results in terms of metrological performance and computing time. The conclusion is that CNNs like StrainNet offer a viable alternative to DIC, especially for real-time applications.
Fichier principal
Vignette du fichier
boukhtache_when.pdf (19.93 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-02933431 , version 1 (08-09-2020)



Seyfeddine Boukhtache, Kamel Abdelouahab, François Berry, Benoît Blaysat, Michel Grediac, et al.. When Deep Learning Meets Digital Image Correlation. Optics and Lasers in Engineering, 2021, 136, pp.106308. ⟨10.1016/j.optlaseng.2020.106308⟩. ⟨hal-02933431⟩
1390 View
213 Download



Gmail Facebook X LinkedIn More