Background: Digital Image Correlation (DIC) is an image-based measurement technique routinely used in experimental mechanics, which provides displacement and strain maps of an observed surface/volume. The metrological performance of DIC has reached its limit which is directly determined by the texture of the imaged surface/volume. Objective : This paper proposes a novel DIC strategy, which relies on a virtual image. This image, noiseless and of infinite resolution, is moreover optimized for providing measurements with the best metrological performance. Methods : The so-called Virtual DIC retrieves the displacement fields by comparing this virtual image to the experimental images. No interpolation is required and processing optimal textures such as checkerboards is possible. Results : Virtual DIC is first applied on synthetic images for comparison purposes with a usual DIC approach. Outstanding metrological performance is observed thanks to the possibility of processing checkerboard patterns. Conclusions : The proposed Virtual DIC is twofold: (i) thanks to the use of a closed-form expression, built-in DIC operators are elaborated without recurring to noisy and poorly defined real images. Interpolation is therefore avoided; (ii) it makes possible it to process checkerboard patterns, which offers the best metrological performance.