A pointwise optimal subset selection strategy assisted by shape functions in digital image correlation algorithm

In the practical application of the digital image correlation (DIC) technique, the local imperfection of the speckle pattern and complex unknown deformation usually make it conflicting to choose an appropriate subset size for each individual sampling point on the tested specimen surface. In this paper, a pointwise shape functions-assisted subset selection (SFSS) strategy for DIC was proposed with the aid of the so-called deformation deviation function. Firstly, a WBSSD correlation criterion was proposed by introducing the Butterworth window function to the classic SSD correlation function, so that the subset size could be continuously adjusted in subpixel level. In addition, the deformation deviation function was taken as an indicator to present the matching degree of the affine shape function in the subset. In this basis, an initial traversing-fine targeting procedure was designed to determine the optimal subset size for each individual sampling point by quantitatively evaluating the defor-mation deviations in a given range of subset sizes. Subsequently, the feasibility and effectiveness of the newly -built SFSS strategy were verified by a series of synthetic speckle images, implying that the presented strategy could enhance the measurement accuracy and precision of DIC algorithm by selecting an appropriate subset size control parameter in WBSSD criterion for each sampling point on the specimen surface. Comparing the proposed SFSS method with the traditional SSD-based correlation algorithm as well as WSSD and WZNSSD algorithms in DIC calculations, it was further demonstrated that the proposed SFSS strategy could improve the measurement performance of DIC, even in the condition of larger complex deformation cases.

Automated summary

In this paper, a pointwise shape functions-assisted subset selection (SFSS) strategy for digital image correlation (DIC) was proposed. The subset size could be continuously adjusted in subpixel level through introducing the Butterworth window function. The feasibility and effectiveness of the SFSS strategy were verified by a series of synthetic speckle images, showing enhanced measurement accuracy and precision in DIC algorithm.