Airship skin strain measurement based on adaptive digital image correlation

Aiming at the non-uniform strain in the welding of airship skin, this paper studies the traditional digital image correlation algorithm, optimizes the shape and size of the subset, and presents an adaptive subset algorithm. Under the premise of ensuring a substantially constant subset area, the modified algorithm uses the initial u and v values to optimize the shape of the subset, find the optimal aspect ratio. Besides, the modified digital image correlation dynamically adjusts the shape of the subset according to the threshold value, making the algorithm better applicable to non-uniform deformation and large deformation. As the results show, in the simulation experiment, under the different strains such as 1000 μɛ, 5000 μɛ, and 9000 μɛ, the standard error measured by the modified method is reduced to more than 50% compared with the traditional method. Meanwhile, during the real speckle experiment, the error in the modified digital image correlation algorithm was reduced from 9.3e−04 to 7.2e−04. Furthermore, the modified digital image correlation only costs 3% more time, has a little effect on computational efficiency. Therefore, the modified method can smooth the measured strain, and make the robustness and accuracy better during the airship skin measurement.