A q-Factor-Based Digital Image Correlation Algorithm (qDIC) for Resolving Finite Deformations with Degenerate Speckle Patterns

Digital image correlation (DIC) has become a widely utilized non-contact, full-field displacement measurement technique for obtaining accurate material kinematics. Despite the significant advances made to date, high resolution reconstruction of finite deformations for images with intrinsically low quality speckle patterns or poor signal-to-noise content has not been fully addressed. In particular, large image distortions imposed by materials undergoing finite deformations create significant challenges for most classical DIC approaches. To address this issue, this paper describes a new open source DIC algorithm (qDIC) that incorporates cross-correlation quality factors (q-factors), which are specifically designed to assess the quality of the reconstructed displacement estimate during the motion reconstruction process. A q-factor provides a robust assessment of the uniqueness and sharpness of the cross-correlation peak, and thus a quantitative estimate of the subset-based displacement measure per given image subset and level of applied deformation. We show that the incorporation of energy- and entropy-based q-factor metrics leads to substantially improved displacement predictions, lower noise floor, and reduced decorrelation even at significant levels of image distortion or poor speckle quality. Furthermore, we show that q-factors can be utilized as a quantitative metric for constructing a hybrid incremental-cumulative displacement correlation scheme for accurately resolving very large homogeneous and inhomogeneous deformations, even in the presence of significant image data loss.