A digital speckle stereo matching algorithm based on epipolar line correction
Liping Liu,Boya Niu,Zhuo Xu,Songyang Zhang,Zhaoyu Shao,Xinyu Wang
DOI: https://doi.org/10.1007/s11760-024-03096-2
IF: 1.583
2024-03-17
Signal Image and Video Processing
Abstract:When the digital speckle correlation method captures images under certain working conditions, the extreme tilt of the camera leads to a weak correlation between the left and right images, which in turn makes the matching difficult. In order to solve the above problem, a digital speckle stereo matching algorithm based on epipolar correction is proposed in this paper, and an estimation method of deformational iterative initial values of the first-order shape function in the stereo matching process is given through theoretical derivation. The Newton–Raphson (NR) method can be directly used for fine matching in the original left and right images through the estimated deformational iterative initial value. The digital speckle matching experiment under different camera inclination angles of 24.2°, 34.2°, 41.15°, 47.3° and 52.4° was carried out to verify the effectiveness of the digital speckle stereo matching algorithm based on epipolar correction. The experimental results show that this method can complete the matching well under different inclined field of view, and the matching success rate is close to 100%, which can at least meet the requirement of digital speckle matching and deformation strain measurement under the inclination angle less than 52.4° field of view.
engineering, electrical & electronic,imaging science & photographic technology