Adaptive Cross Correlation for Imaging Displacements in Soils

Digital image correlation (DIC) is used in this paper to study two-dimensional spatial soil deformations nonintrusively. Adaptive cross correlation (ACC), which is an advanced cross-correlation algorithm that utilizes variable window sizing and window shifting methods, was used to reduce the errors associated with conventional DIC. Comparison of the two algorithms using a scheme of predefined digital and physical movements shows that ACC improves the accuracy and range of DIC. A model of a strip footing on sand is presented. A digital camera was used to capture consecutive images of soil deformations through a Plexiglas mold. The spatial deformation of the soil beneath the footing was obtained by correlating consecutive images using ACC. Shear strains and the failure surface were also calculated based on the displacements measured using ACC. The optical setup and digital image processing technique presented in this paper were proven to be an effective tool for studying soil–structure interaction nonintrusively.