Development of the 3D digital image correlation technique

• Main Author: Qiu, Lili
• Other Authors: Qian Kemao
• Format: Final Year Project
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/62882
• Institution: Nanyang Technological University

The image correlation has been extensively studied over the decades and became increasingly popular in the field of experimental solid mechanics to obtain the strain measurement. The two dimensional correlation manages relatively small strains by comparing images before and after deformation. And it is not able to produce accurate result when surface deformation happen. With the limitation of 2D digital image correlation, there is a need to explore the 3D DIC technique to produce accurate result in three dimensional space. In this work, the 3D correlation technique is developed to quantify the displacement in three dimensional space by incorporating Fourier transform cross correlation to obtain initial estimation of the displacement and inverse compositional Gauss-Newton (ICGN) with pre-computed coefficient of Bicubic interpolation and hessian matrix to accelerate the 2D correlation computation process. One reference image is chosen to correlate with another three images capture before and after deformation. The stereo vision system is employed to reconstruct the 3D coordinate based on intrinsic and extrinsic parameters of left and right camera, after which the 3D displacement is calculated. In addition, the performance of the 3D correlation is evaluated in term of accuracy and speed.