Smoothing displacement field obtained via digital image correlation for robust strain calculation

Abstract

A self-adaptive smoothing method is proposed to deal with the displacement field obtained via digital image correlation (DIC). This new smoothing method is based on penalized least squares regression technique, which can penalize the roughness of the displacement and realize displacement denoising. In this method, generalized cross validation (GCV) method and discrete cosine transform (DCT) are used to get penalized parameter and true displacement from the noisy displacement field separately. This method has advantages of easy implementation, little calculation requirement, complete automation and so on. Both simulation analyses and experimental results validate the effectiveness of this smoothing technology, which can improve the strain distribution differentiated by the smoothed displacement without any calculation increasement.