基于数字图像相关法在建筑结构变形场中的抗噪研究及应用

(1.西安建筑科技大学 信息与控制工程学院,陕西 西安 710055; 2.西安建筑科技大学 土木工程学院,陕西 西安 710055)

数字图像相关法; 混凝土; 抗噪

Research and application of anti-noise in deformation field of building structure based on digital image correlation method
ZHANG Wu1, ZHAO Liang1,2, XU Lin1,WANG Fan1, LU Ying1

(1.College of Information and Control Engineering, Xi'an Univ. of Arch. & Tech., Xi'an 710055, China; 2.School of Civil Engineering, Xi'an Univ. of Arch. & Tech., Xi'an 710055, China)

digital image; concrete; anti-noise

DOI: 10.15986/j.1006-7930.2021.05.015

备注

对建筑结构的变形进行测量时存在操作繁琐、易受外界影响等问题,采用数字图像相关法可以克服这些影响,但是,由于实际环境产生的噪声会影响测量结果的准确性,进而影响建筑结构变形场的分析.本文提出一种新的数字图像相关法,利用Huber函数正则化来得到建筑结构图像的灰度梯度,然后结合N-R亚像素配准算法得到亚像素位移,来提高算法的抗噪能力和稳定性.利用仿真实验验证了该方法的优越性和可靠性,并制备了混凝土试件,采用新方法获取混凝土单轴压缩下的应变场并分析了单轴压缩下试件裂纹起裂、扩展及破坏模式.结果表明:新方法具有良好的抗噪能力,可以直观地以图像的形式展现混凝土试件表面的裂纹变化,从细观角度为研究混凝土表面的裂纹演化规律和损伤特征提供了稳定有效的可视化技术.
When measuring the deformation of building structure, there are problems such as cumbersome operation and susceptibility to external influences. The digital image correlation method can overcome these effects, but the noise generated by the actual environment will affect the accuracy of the measurement results, and then affect the deformation field of the building structure analysis. This paper proposes a new digital image correlation method, which uses Huber function regularization to obtain the gray gradient of the architectural structure image, and then applies the N-R sub-pixel registration algorithm to obtain the sub-pixel displacement to improve the anti-noise ability and stability of the algorithm. The superiority and reliability of the proposed method were verified by simulation experiments, and concrete specimens were prepared. The strain field of concrete specimens under uniaxial compression was obtained by using the new method, and the crack initiation, propagation and failure modes of specimens under uniaxial compression were analyzed. The results show that the new method has good anti-noise ability and can visually show the crack changes on the surface of concrete specimens in the form of images. It provides a stable and effective visualization technique for studying the crack evolution law and damage characteristics of the concrete surface from a microscopic point of view.