双目视觉下振动台试件变形位移测量方法研究

发布时间：2018-06-05 01:05

本文选题：机器视觉技术 + 人工标志　；参考：《兰州理工大学》2017年硕士论文

【摘要】：振动台试验是研究结构试件抗震性能的一种重要方法,通过对振动台输入不同的激励信号,获取结构试件的特定测试点在不同激励信号下的速度、位移等运动参数。其中对特定测试点的位移进行相关分析,得到特定测试点处结构试件的变形位移具有重要的实际意义。本文将机器视觉技术应用到振动台试验中结构试件特定测试点处变形位移测量中,利用两台摄像机同步对随振动台一起运动的结构试件进行视频采样,然后通过对图像进行相关处理与分析,基于双目视觉测量原理及本文提出的测量算法计算出结构试件的变形位移,从一种新的渠道为振动台试验提供更广泛的数据。本文研究的主要工作内容有以下几个方面:(1)为了提高结构试件变形位移的测量精度,设计了一种用回光反射材料制作的人工圆形标志粘贴在被测结构试件表面,利用图像处理相关知识对粘贴有人工圆形标志的试件图像进行处理分析,通过对几种处理方法的处理效果及处理速度的对比,从中选择出相应的合适算法。针对处理过程中出现人工标志与背景粘连进行了处理,消除了粘连,并利用改进的最小二乘椭圆拟合算法对人工标志进行了拟合定位。(2)针对左右两摄像机对应时刻图像中人工标志点的匹配问题,采用了基于对称性测验的极线约束准则。对于序列图像中人工标志点的跟踪,采用了Kalman滤波算法。(3)对摄像机标定原理进行了深入的研究分析,并利用张氏标定法对摄像机进行了标定,在标定实验中不但考虑了畸变,而且对精度采取多层次细化纠错,完成了对摄像机的高精度标定。另外针对本文的研究对象,在双目视觉测量原理基础上提出了自己的测量算法,并对振动台在固定位移信号下试件的位移和正弦波信号激励下试件的变形位移进行了测量,由测量结果可知符合客观事实。
[Abstract]:Shaking table test is an important method to study the seismic behavior of structural specimens. By inputting different excitation signals to the shaking table, the motion parameters such as velocity and displacement of specific test points of structural specimens under different excitation signals are obtained. Among them, it is of great practical significance to analyze the displacement of specific test points, and to obtain the deformation displacement of structural specimens at specific test points. In this paper, the machine vision technology is applied to the deformation displacement measurement at the specific test point of the structural specimen in the shaking table test. Two cameras are used to synchronize the video sampling of the structural specimen moving with the shaking table. Based on the principle of binocular vision measurement and the measurement algorithm proposed in this paper, the deformation displacement of the structural specimen is calculated, which provides more extensive data for the shaking table test from a new channel. In order to improve the measuring accuracy of deformation displacement of structural specimens, a kind of artificial circular mark made of backlight reflection material has been designed and pasted on the surface of tested structural specimens. The image with artificial circular sign is processed and analyzed by using the relevant knowledge of image processing. By comparing the processing effect and processing speed of several processing methods, the appropriate algorithm is selected. In view of the artificial marks and background adhesions occurred in the process of processing, the adhesions were eliminated. The improved least square ellipse fitting algorithm is used to locate the artificial mark. (2) aiming at the matching problem of the artificial mark points in the image of the left and right cameras corresponding to the time, the pole line constraint criterion based on symmetry test is adopted. For the tracking of artificial mark points in the sequence image, the principle of camera calibration is deeply studied and analyzed by using Kalman filter algorithm. The camera calibration method is used to calibrate the camera. The distortion is not only considered in the calibration experiment, but also discussed in this paper. Furthermore, the precision of the camera is calibrated by multi-level refinement and error correction. In addition, based on the binocular vision measurement principle, the measurement algorithm is put forward, and the displacement of the specimen under the fixed displacement signal and the deformation displacement of the specimen excited by the sine wave signal are measured. From the measurement results, we can see the objective facts.
【学位授予单位】：兰州理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TB534.2

【参考文献】