Markers tracking and extracting structural vibration utilizing Randomized Hough transform

摘要

The structural vibrations data may be used for health-monitoring of structures, determination of the structural characteristics, or up-date the structural analysis model. Therefore, accuracy in the measurement of these vibrations is crucial. Extracting the structural vibrations by utilizing accelerometers or LVDT1 sensors has different limitations, such as wiring restrictions, accessibility, and sometimes a change in the mass and stiffness of the structure. In recent years, we have witnessed an increase in the use of remote sensing. One of these methods is the use of camera and image processing. In this paper, a new algorithm is used to track marker points, which are attached to the object to extract structural dynamic characteristics like mode shapes and natural frequencies. For this purpose, the Randomized Hough detection algorithm is used to observe the ellipse markers movement on the structure. So, a three-story frame was considered and the vibration, mode shape and the natural frequency were derived from the new algorithm. The results have been compared with other measurement techniques like the Lucas-Kanade algorithm, the accelerometer sensors, and finite element method. The consequences indicate that the new algorithm improves computer vision in tracking points for extracting structural dynamic properties.