Mapping Motion-Magnified Videos to Operating Deflection Shape Vectors Using Particle Filters
Phase-based motion estimation and magnification are targetless methods that have been used recently to perform experimental modal analysis (EMA) and operational modal analysis (OMA) on a variety of structures. Mapping the motion-magnified sequence of imag
- PDF / 899,619 Bytes
- 9 Pages / 595.516 x 790.987 pts Page_size
- 74 Downloads / 135 Views
Mapping Motion-Magnified Videos to Operating Deflection Shape Vectors Using Particle Filters Aral Sarrafi and Zhu Mao
Abstract Phase-based motion estimation and magnification are targetless methods that have been used recently to perform experimental modal analysis (EMA) and operational modal analysis (OMA) on a variety of structures. Mapping the motionmagnified sequence of images into quantified operating deflection shape (ODS) vectors is currently being conducted via edge detection methods that require intensive human supervision and interference. Within this study, a new hybrid computer vision approach is introduced to extract the quantified ODS vectors from the motion-magnified sequence of images with minimal human supervision. The particle filter point tracking method is utilized to follow the desired feature points in the motion-magnified sequence of images. Moreover, the k-means clustering algorithm is employed as an unsupervised learning approach to performing the segmentation of the particles and assigning them to specific feature points in the in the motionmagnified sequence of images. This study shows that the cluster centers can be employed to estimate the ODS vectors, and the performance of the proposed methodology is evaluated experimentally on a lab-scale cantilever beam and validated via a finite element model. Keywords Phase-based motion estimation · Video magnification · Particle filter · Computer vision · Clustering · Unsupervised learning
8.1 Introduction In recent years using digital cameras for structural dynamics identification and structural health monitoring (SHM) has been gaining much attention [1, 2]. As a non-contact measurement method, digital cameras have several advantages over other traditional approaches for structural dynamics identification. In general, instrumentation of structures with accelerometers is labor-intensive, time-consuming and costly, but for digital cameras, the instrumentation is normally much easier and attaching a couple of optical targets is sufficient in most of the cases. Another major advantage of camera-based measurement over traditional structural response measurement methods such as accelerometers or laser vibrometers is that the cameras can provide full-field measurements in the field of view of the camera simultaneously [3]. Laser vibrometers can provide high spatial resolutions [4], but normally scanning the surface of the measurement will take a long time that may increase the testing time dramatically. Although cameras-based measurements have several advantages compared to other techniques, the post-processing of the sequence of images captured by the camera is generally an elaborate and complex task which involves advanced signal processing [5, 6], computer vision, and computational photogrammetry algorithms. Computer vision is one of the fastest growing branches in computer science that aims to extract useful information from the images captured from a scene. 3-dimensional digital images correlation (3D-DIC), and 3-dimensional point tracking (3DPT
Data Loading...
