27 May 2023 in Optics Communications
22 December 2023 in IEEE Sensors Journal
Transverse vibration measurement of beam structures based on computer vision has been widely used, but most visual measurement methods require the installation of artificial markers at the measured positions. This makes the measurement of vibrations in some key position of beam structures difficult due to their inaccessibility. Therefore, the development of target-less approaches is important for some engineering applications. In this paper, a method for measuring displacement and vibration of beam structures visually is presented based on sub-pixel edge point tracking. Combining the principle of coordinate space transformation and the two-dimensional normal distribution characteristics of false edge parameters, false edges are eliminated to determine the position of structural edges accurately. This allows measurement of beam deflection and point tracking of the beam structure’s vibration. The effectiveness of the proposed method is verified through a cantilever beam vibration experiment and the measured data are compared using a traditional piezoelectric sensor (eddy current displacement sensor and accelerometer) and digital image correlation method. Compared with the traditional piezoelectric sensor methods, the maximum difference of the measured displacement was less than 2.4%. Compared with the digital image correlation method, the time series correlation of the proposed method with the reference measurements was 0.35% higher, and the main vibration frequency error was 0. The experiment proves that the proposed method can be used as an effective visual measurement method for target-less tracking of the transverse vibration of beam structures.
4 July 2023 in IEEE Access
During camera calibration, the plane target image is required to occupy most of the whole image, and the focused image is also needed. It is difficult to satisfy these conditions in large-scale structural vision measurement. Therefore, it is very important to explore a camera calibration method with a small calibration board for a large field of view. This paper proposed an out-of-focus image calibration method based on the accurate positioning of concentric circular projection center, which is not limited by the camera’s field of view and can overcome the influence of out-of-focus image on camera calibration. The proposed method consists of three parts: adaptive blurred circle image enhancement, accurate positioning of concentric circle projection center and camera calibration of concentric circle center point array. The simulation results show that under different Gaussian blur kernel radius conditions, the proposed accurate positioning method of the concentric circle projection center has better robustness. When Gaussian kernel radius K=40, the positioning accuracy of the proposed method is improved by 41.88%.In the real image experiment, the internal parameters of the camera in the heavy defocus image have similar results to the Zhang’s calibration method in the focused image. Both simulation and real image experiments show that the proposed method can realize the calibration of large targets with a small plane calibration board.
15 September 2026 in SSRN Electronic Journal
Vision-based structural vibration measurement encounters two major challenges under dynamic conditions: feature point mismatching and positional inconsistency between consecutive frames. This study proposes a feature-point-correction-based image tracking method to overcome these issues. By utilizing the short-term rigid-body transformation characteristics of vibrating structures, a transformation model for feature point sets is established. The maximum inter-class variance method is employed to adaptively identify mismatched points, and the optimal affine transformation matrix is estimated by minimizing the reprojection error, enabling the recovery and correction of erroneous and inconsistent points. Finally, structural displacements are computed from the centroids of the corrected point sets. Simulation results demonstrate that the proposed method improves centroid localization accuracy by more than 80%. In the column vibration experiment, compared with the eddy-current displacement sensor, the root-mean-square error of displacement is reduced by 67.46%, with a relative deviation of 2.39% and a correlation coefficient exceeding 0.9; the first three natural frequencies are measured with deviations less than 0.87%. In the engineering application on a parachute tower, the relative deviation of peak-to-peak displacement during the ascending phase is 3.05% with a correlation coefficient of 0.90, and 2.50% with a correlation coefficient of 0.89 during the descending phase; the measured vibration displacements are far below the allowable elastic deformation specified by the national standard. The experimental and engineering results consistently demonstrate that the proposed method significantly improves the accuracy and stability of vision-based tracking under dynamic conditions, offering a reliable solution for vibration monitoring of large-scale structures with sparse textures.