Fourier-based image shift for periodic effect compensation in displacement monitoring using vision systems

In a wide range of monitoring applications, vision systems are nowadays applied. The main advantages of vision-based monitoring include the possibility to: i) measure a wide region of the target, obtaining a dense measurement, ii) obtain a 3D estimation of shape, displacements and strain field, iii) measure with a completely contactless technique, therefore avoiding loading effect and wear of measuring components and target. Digital Image Correlation (DIC) is an image processing technique that allows measuring the displacement and strain fields of target, as well as the 3D shape of the measurand in the case of stereo DIC. Displacement measurement obtained by DIC are affected by a systematic effect in displacement data. In this paper, we propose and qualify an image pre-processing technique, to be applied to the images before running the DIC analysis, capable to strongly reduce the systematic effect affecting the displacement results.