Abstract This work evaluates the effect of illumination fluctuations on three-dimensional Digital Image Correlation (3D DIC) measurements that can be performed for the health monitoring of large infrastructures. Usually, correlation criteria are robust against uniform light variations. However, discrepancies may arise when brightness variations affect different regions of the image. Two shadow compensation techniques, derived from pre-existing image enhancement algorithms, are proposed to improve the accuracy of the DIC results. Their effectiveness is first tested on digitally modified images, in which irregular brightness changes are introduced. Next, an outdoor test is conducted, with shadows intentionally applied to part of the monitored region. The brightness adjustment methods moderately reduce the standard deviation of the error, although the mean values are not affected. In addition, the compensation techniques enhance quality metrics, such as the Mean Intensity Gradient (MIG). The improvement in results is most evident in the case when partial shadow is present in the reference image.
Hajjar et al. (Tue,) studied this question.