...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Mechanical systems and signal processing >Measurement of full-field displacement time history of a vibrating continuous edge from video
【24h】

Measurement of full-field displacement time history of a vibrating continuous edge from video

机译：从视频中测量振动连续边缘的全场位移时间历史

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Video of a vibrating structural component provides dense continuous spatial information of motion compared to contact based sensors which are more localized. In order to harness such quantitative information, various Computer Vision based methods have been studied in the past few years. Computer vision algorithms that implement sparse optical flow, tracks the displacement of few selected texture rich patches of pixels corresponding to salient points of the structure, in subsequent frames of the video starting from the initial frame. The obtained displacement response of individual, spatially discrete, points of the structure denotes the Lagrangian representation of motion. Such vision based measurement already proved its potential in replacing the need for a contact based vibration measuring sensors. However, full-field vibration response of a structural component, in its operating condition, is vital for identifying heterogeneous mechanical properties as well as localize damages. In order to obtain full-field, spatially dense, vibration response, a large number of sensors are needed to be attached throughout the specimen's surface, making it impractical. In such scenarios, the technique of Digital Image Correlation (DIC) is used to measure Lagrangian representation of motion using Computer Vision algorithms which implements dense optical flow. In DIC, the presence of much required rich optical textures is ensured by adding speckle paint on the surface of the specimen which can be easily tracked over subsequent frames of the video. But its application is limited to laboratory setups that cannot be implemented for full-scale real-world structural components that vibrate in its operating condition. Continuous edge of a moving object is a rich optical feature whose motion perpendicular to its orientation can be tracked in Lagrangian coordinates using optical flow. In this paper, a novel method of measuring full-field displacement response of a vibrating continuous edge of a structural component is proposed from its video. The proposed method computes both horizontal and vertical components of the Lagrangian displacement of each pixel of the edge for every frame in the video with sub-pixel accuracy. The method consists of applying d'Alembertian of Gaussian filter on the spatiotemporal video signal to convert the edge information into zero crossing signal which is tracked over time using the optical flow of edge. Intermittent and sparse noise in the obtained full-field displacement signal is removed using Robust Principal Component Analysis. Further, experimental validation of the proposed method is presented for two kinds of structures, one a cantilever beam, another a two-story steel frame, both undergoing harmonic as well as random base excitation. The results obtained using the proposed method is validated with the displacement measured using Laser Doppler Vibrometer at a particular point of the edge. Further, the analytical derivations of the proposed method along with the effect of the filter parameters on the edge signal are presented in the Appendix. In order to validate the efficacy of the method on real-world structures, part of the vibrating cable of Tacoma Narrows bridge is tracked from its video, moments before its collapse. The results show high correspondence between the actual motion of the cable and traced motion of its edge, over time. The proposed method reveals the absolute as well as relative displacement response of the cable along with the dominating frequencies of the cable's vibration.
机译：与基于接触的传感器相比，振动结构部件的视频提供了运动的密集连续空间信息。为了利用这种定量信息，在过去几年中已经研究了各种基于计算机视觉的方法。实现稀疏光流量的计算机视觉算法，跟踪少数选择纹理的位移对应于结构的突出点的像素，在从初始帧开始的后续帧中。所获得的个体的位移响应，空间离散，结构点表示拉格朗日的运动表示。这种基于视觉的测量已经证明其潜力在更换基于接触的振动测量传感器时的潜力。然而，在其运行条件下，结构部件的全场振动响应对于识别异质机械性能以及本地化损害至关重要。为了获得全场，空间密集，振动响应，需要大量传感器在整个样本的表面上安装，使其不切实际。在这种情况下，数字图像相关（DIC）的技术用于使用计算机视觉算法来测量运动的拉格朗日表示，该计算机视觉算法实现致密光学流。在DIC中，通过在样本表面上添加散斑涂料来确保需要多次富光学纹理的存在，这可以在视频的后续帧上容易地跟踪。但其申请仅限于实验室设置，无法用于以其运行条件振动的全规模实际结构组件。移动物体的连续边缘是一种丰富的光学特征，可以使用光学流动在拉格朗日坐标中跟踪垂直于其取向的运动。本文从视频中提出了一种测量结构部件的振动连续边缘的全场位移响应的新方法。所提出的方法为视频中的每个帧具有子像素精度来计算Lagrangian的每个像素的水平和垂直分量。该方法包括在时空视频信号上应用D'Anementian在时空视频信号上，将边缘信息转换为零交叉信号，其使用边缘的光学流动随时间跟踪。使用鲁棒主成分分析除去所获得的全场位移信号中的间歇和稀疏噪声。此外，提出了这两种结构的实验验证，介绍了两种结构，一个悬臂梁，另一个是一个两层钢框架，都接受谐波以及随机基础励磁。使用所提出的方法获得的结果通过在边缘特定点使用激光多普勒振动计测量的位移。此外，所提出的方法的分析推导以及滤波器参数对边缘信号的效果呈现在附录中。为了验证对现实世界结构的方法的功效，从其崩溃之前的视频，塔科马狭窄桥的振动电缆的一部分振动电缆。结果显示了电缆的实际运动与其边缘的追踪运动之间的高对应关系。所提出的方法揭示了电缆的绝对和相对位移响应以及电缆振动的主导频率。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号