Wing deformation measurement using the stereo-vision methods in the presence of camera movements

2021 
Abstract The stereo-vision methods have been developed and used for measuring the wing deformation while the aircraft is in flight. However, it is difficult to achieve accurate measurement in flight test since inevitable camera movements often occur due to elasticity in the camera support and the surrounding structure, especially during heavy turbulence and large maneuver flight. Any slight camera movements within a stereo-vision system can directly influence the calibrated extrinsic parameters, then decrease 3D reconstruction accuracy, and even cause the measurement to fail. This study proposes a method to correct the effect of the camera movements to enable accurate wing deformation measurement using a stereo-vision system. A reference coordinate system is established based on the control points to unify the changing camera coordinate system. A dynamic extrinsic parameters calibration algorithm is developed to estimate the instantaneous position and orientation of the stereo cameras. To verify the presented method, an experiment in which the stereo cameras occurring relative movement was performed. The results show that the proposed method has effectively corrected the effects of camera movements. Furthermore, the proposed method was applied to the practical wing deformation measurement of a UAV in fieldwork. The maximum error of length and displacement is less than 0.8 mm when the stereo cameras remained unchanged. The actual accuracy of the overall measurement system is about 0.01% of the observed area (e.g. 0.8 mm in 6 m scale) under the conditions of flat viewing angle and outdoor scenes. While camera movements occur during the test, the corrected RMS error of displacement is reduced to about 1.5 mm and has decreased by 50% compared with that before correction. The enhanced precision and robustness of wing deformation measurement are well realized by the proposed method of correcting camera movements.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    41
    References
    0
    Citations
    NaN
    KQI
    []