雷振坤
点击次数：
论文类型：期刊论文
发表时间：2017-03-01
发表刊物：OPTICS AND LASERS IN ENGINEERING
收录刊物：SCIE、EI
文献类型：J
卷号：90
页面范围：48-58
ISSN号：0143-8166
关键字：Digital image correlation; Inverse compositional matching algorithm; Invertible second-order shape operator; Large deformation measurement
摘要：Compared with the traditional forward compositional matching strategy, the inverse compositional matching strategy has almost the same accuracy, but has an obviously higher efficiency than the former in digital image correlation (DIC) algorithms. Based on the inverse compositional matching strategy and the auxiliary displacement functions, a more accurate inverse compositional Gauss-Newton (IC-GN2) algorithm with a new second-order shape operator is proposed for nonuniform and large deformation measurements. A theoretical deduction showed that the new proposed second-order shape operator is invertible and can steadily attain second-order precision. The result of the numerical simulation showed that the matching accuracy of the new IC-GN2 algorithm is the same as that of the forward compositional Gauss-Newton (FC-GN2) algorithm and is relatively better than in IC-GN2 algorithm. Finally, a rubber tension experiment with a large deformation of 27% was performed to validate the feasibility of the proposed algorithm.
