Structural Dynamics Identification Using Motion Estimation and Video Magnification

该项目旨在探索使用一种基于摄像机的非接触式方法来识别结构的微小运动和动态行为。这有可能帮助工程师了解3D结构的复杂动态行为，例如，包括风力涡轮机、桥梁、飞机、船舶和汽车。这将有助于结构健康监测，以及测试大型复杂结构变形的数值模型预测。可以使用为手机开发的便携式电子设备应用程序(APP)来记录视频、提取变形和动态行为。这将允许研究人员、设计工程师和现场检查员简化对该技术的使用。研究团队将与高中生合作，将视频传感技术整合到他们的课程中。该项目还将涉及女性和少数民族学生。该项目的主要研究目标是了解和量化通过摄像机视频运动放大提取的结构动力学与真实动态运动之间的关系。该项目有可能改变目前进行大规模实验模态分析和三维(3D)结构动力学识别的方式。这项工作的成功完成将有助于更好地理解基于视觉的结构动态参数提取以及处理后的数据如何受到环境和操作变量的影响。这项工作将把基于相位的运动估计技术与3D数字图像相关和点跟踪相结合，以实现对如何使用二维相机测量来提取定量3D运动的理解。该项目的成功成果将解决影响相关算法的因素与处理结果之间的现有知识差距，同时量化基于视频的结构动力学识别的不确定性。这项研究将产生对提取的光流如何与动态运动相关联的系统理解。定量了解不同参数如何影响基于相位的运动提取和视频放大的结果，将导致一种变革性的非接触测量新方法，该方法可能会影响各种基础设施系统的系统识别、验证和结构健康监控。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This project aims to explore the use of a non-contact, video camera-based method to identify small motions and dynamic behavior of structures. This has the potential to help engineers understand complex dynamic behavior of 3D structures, including, for example, wind turbines, bridges, aircraft, ships, and automobiles. This will help in structural health monitoring, as well as in testing numerical model predictions of deformations of large, complex structures. Videos can be recorded, and deformations and dynamic behavior distilled using a portable electronic device application (app) developed for a cell phone. This will permit simplified use of the technology by researchers, design engineers and field inspectors. The research team will work with high school students to integrate the video sensing technology into their curriculum. The project will also involve female and minority students.The primary research objective of this project is to understand and quantify the relationship between the structural dynamics extracted via camera video motion magnification and the true dynamic motion. The project has the potential to transform the way large-scale experimental modal analysis and three-dimensional (3D) structural dynamics identification is currently performed. The successful completion of the work will lead to a better understanding of vision-based structural dynamic parameter extraction and how the processed data is affected by environmental and operational variabilities. This work will integrate the phase-based motion estimation technique with 3D digital image correlation and point tracking to achieve an understanding of how two-dimensional camera measurements can be used to extract quantitative 3D motion. The successful outcomes of this project will address the existing knowledge gap between the factors that contaminate the related algorithms and the processed results, while quantifying uncertainty for video-based structural dynamics identification. The research will generate a systematic understanding of how extracted optical flows are associated with dynamic motions. Achieving a quantitative understanding of how different parameters influence the results for phase-based motion extraction and video magnification will lead to a transformative new approach for non-contact measurement that can impact system identification, validation, and structural health monitoring of various infrastructure systems.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.