Subpixel In-Plane Displacement Measurement Using Digital Image Correlation and Artificial Neural Networks

Development in digital image correlation have made it a widely and effective tool for full-field displacement and displacement gradients measurement in solid and fluid mechanics. The main method to improve accuracy is to use the subpixel registration algorithm in digital image correlation. This paper presents an in-plane subpixel displacement detection method based on Fourier Transformation and ANN. Correlation coefficients of subimages in undeformed and deformed images of specimen surface have a peak, the position of which represent the integer pixel displacement. Subpixel displacement is obtained in this paper by training ANN to estimate. Results from computer-simulated images indicate techniques can obtain similar accuracies compared with other subpixel algorithms, but the ANN approach has the advantage that it can complete very fast subpixel displacement analysis without knowledge of the analytical form of local correlation coefficient.