Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
729563 | Measurement | 2015 | 12 Pages |
•This paper proposes a stereovision-based 3D deformation measurement method.•A Canny-Zernike combination algorithm achieve 0.02 accuracy for edge detection.•Kalman filter is used to track circular targets on dynamic sequential images.•A free vibration test of a stayed-cable model is conducted in lab.•Experiment results verify the reliability and feasibility of the proposed method.
To monitor 3D deformations of structural vibration response, a stereovision-based 3D deformation measurement method is presented in this paper. The center extraction and feature matching algorithm of circular target is studied to obtain 3D displacement of structural vibration response. Compared with traditional edge detection algorithms, this method utilizes a Canny-Zernike combination algorithm to attain sub-pixel edge of circular target, and the location accuracy of circular target can achieve 0.02 pixel. A free vibration test of a stayed-cable model is conducted in lab. The out-of-plane and in-plane displacement time histories of three circular targets on the model are gained simultaneously. Experiment results indicate the proposed method has the good agreement with the laser LVDT in in-plane data, the first three order frequencies of stayed-cable model identified by two methods are almost consistent. It is verified that this proposed approach is reliable and useful for monitoring 3D deformations of structural vibration response.