Weak Texture Enhancing Method Based on Curvelet Transformation and γ-CLAHE Correction in Digital Image Correlation
摘要
Weak texture images may easily reduce the quality of images, which affects the measurement accuracy of digital image correlation (DIC) test. Although several methods in the filed of 3D surface reconstruction and features extraction of weak texture images had been proposed, few of them focused on the displacement and deformation measurement of DIC, and the effects of weak texture after being processed on the DIC test have not been considered. To solve this problem, a weak texture enhancing method is proposed in this study, which is conducted by decomposing the weak texture image into curvelet domain, and filtered to a representative texture scale-image based on Gray Level Concurrence Matrix (GLCM) and mean intensity gradient (MIG), then processing the representative texture scale-image using γ-correction and contrast limited adaptive histogram equalization (CLAHE). Through the numerical uniaxial displacement and deformation experiment, physical rigid body in-plane rotation test and four-point bending test of weak texture images, the measurement performance of this proposed method is validated. The results show that the proposed method can efficiently optimize the speckle qualities of the weak texture images, and markedly improve the measurement accuracy in DIC test.