DIC measurement of large-scale objects based on global–local optimization image stitching

Digital Image Correlation (DIC)is a non-contact measurement method based on computer vision. It achieves displacement and strain measurement by matching the speckle pattern of the same pixel in the reference and target subsets before and after deformation. As the application fields of DIC continue to expand, the test pieces are becoming more and more diversified. Therefore, this paper proposes a global–local optimization image stitching method to achieve full-field measurement of large-scale objects. First, speckle images with overlapping parts are collected at different locations of the objects. Secondly, the collected images are stitched through the global–local optimization algorithm to obtain a panorama image. Finally, the DIC algorithm is applied to calculate the whole deformation field. In the experiments, we verify the feasibility and noise immunity of the proposed image stitching method through numerical simulation, and this method has smaller measurement errors than the conventional Scale-Invariant Feature Transform (SIFT) and Speeded Up Robust Features (SURF) algorithms. Then, the accuracy of this method in practical applications is verified by tensile and three-point bending experiments. The experimental results show that the method proposed in this paper extends the application range of the DIC method.