Displacement Measurement of Shaking Table Experiments Based on Binocular Vision and Subpixel Optimization

Abstract

The shaking table experiment is a very useful tool to assess the seismic performance of structural models. However, there are many shortcomings in traditional displacement measurement methods. In this research, a binocular vision algorithm for accurately extracting the subpixel center of the circular target based on elliptic fitting is proposed. Firstly, the images are collected by two high-speed cameras and pre-processed to enhance their brightness and contrast. Then, an improved subpixel optimization method based on the Zernike moments and Canny operator is adopted to get accurate localization of extracted target points’ three-dimension coordinates. After comparing the displacement of target points with laser sensors, the results show that the proposed algorithm achieves high-precision displacement measurements and overcomes the problems that traditional methods hold.