Random errors in DIC caused by non-uniform image noise

Digital image correlation (DIC) is an optical measurement technique widely used in the field of experimental mechanics due to its practicality, flexibility, and reliability. The principle of DIC is to correlate the same point in the reference and deformed image. The performance of DIC is influenced by the image noise. Theoretical models have been presented to evaluate the random errors caused by image noise. However, these models are based on the assumption that the variance of image noise is uniform, while in practice the variance of image noise is non-uniform. In order to overcome this deficiency, in this paper, a formula of random errors of DIC corresponding to non-uniform noise is derived. The formula shows that the variance of image noise and image gradients has a coupling effect. In order to verify our theoretical analysis, actual experiments were carried out. The dependence between image noise and intensity was measured for two cameras; the DIC errors caused by image noise were measured experimentally; in order to take the effect of illumination into account, we measured the DIC errors in cases of both uniform and non-uniform illumination. The experimental results show good agreements with our theoretical analysis.