Monocular out-of-plane vibration measurement using parametric homography

Vision-based measurement is an emerging method for vibration tests due to its non-contact nature and high spatial resolution, which make the full-field measurement possible. Conven-tional vision-based measurement requires two or more cameras to produce an out-of-plane response measurement. However, the additional cost of another camera and synchronization system makes applying this kind of measurement method to real-world engineering projects more challenging. This paper presents a single-camera method for measuring the out-of-plane vibration of plate structures. To accomplish that, the image variation induced by the motion of a plane is represented using a motion-parametric homography. Under the Lucas-Kanade framework, the motion-parametric homographies are estimated separately for each region of interest (ROI) between images, and the out-of-plane displacement response is obtained concurrently. The proposed method's results are confirmed in good agreement with those of the laser Doppler vibrometer through an impact test on a steel plate structure.

Automated summary

Vision-based measurement is an emerging method that enables full-field measurement with non-contact and high spatial resolution capabilities. This paper presents a single-camera method for measuring out-of-plane vibration of plate structures using motion-parametric homography to capture image variation and displacement response.