Digital volume correlation in an environment with intensive salt-and-pepper noise and strong monotonic nonlinear distortion of light intensity

Digital volume correlation is an image-based technique for internal 3D displacement and strain fields measurement or analysis widely used in the field of experimental mechanics. A widely used correlation function (criterion) of digital volume correlation is Pearson correlation function, which suffers from the problem of the acquired data being contaminated by salt-and-pepper noise and monotonic nonlinear distortion of the light intensity. In this work, a 3D correlation function called the Spearman correlation function is used to deal with those interferences. A numerical experiment shows that the performance of Spearman correlation function using integer-pixel registration in an environment with 10% salt-and-pepper noise is better than that of Spearman and Pearson correlation functions using sub-pixel registration in an environment with 1% salt-and-pepper noise. As the light intensity distortion is significant, the error of Pearson correlation function is considerable; meanwhile, the error of Spearman correlation function is small. In conclusion, Spearman correlation function is, in particular, practical and useful in digital volume correlation.

---

Opracowanie ze środków MNiSW w ramach umowy 812/P-DUN/2016 na działalność upowszechniającą naukę (zadania 2017).