Abstract: Combined with micro-CT in-situ loading scanning, a digital volume correlation was used to identify and quantify sub-voxel level micro-cracks in a mortar sample. The three-dimensional volumetric images of mortar specimens under different uniaxial compression stages were acquired by using a micro-CT in-situ loading system, and the failure characteristics were analyzed. The regularized global digital volume correlation and the principle of micro-crack identification were presented, and then the measurement precision and micro-crack identification were evaluated and analyzed. Based on the 3D displacement fields, the damaged elements were identified in consideration of gray level residual and maximum principal strain, and then the damage factor was defined to identify and to quantify the evolution of nascent micro-cracks. The results show that the introductions of the regularization and damage factor improve the measurement accuracy of the global digital volume correlation and detect nascent sub-voxel level micro-fractures. The fine displacement fields and strain fields clearly depict the localized deformation characteristics around pores and fractures in the sample. In the local region of the sample, subvoxel-level micro-fractures occur in the elastic stage, and the proportions of nascent subvoxel-level micro-fractures in the plastic stage and the early stage of destruction are higher than those of voxel-level micro-fractures. This study provides both a new idea and a method for the identification and quantification of micro-cracks in concrete.