Shu Xiaosong, Bao Tengfei, Hu Yuhan, Li Yangtao, Zhang Kang
Appl Opt. 2021 Dec 1;60(34):10477-10489. doi: 10.1364/AO.444593.
Camera calibration is essential for various vision-based 3D metrological techniques. In this paper, a novel camera calibration method, to the best of our knowledge, combining synthetic speckle pattern and an improved gray wolf optimizer algorithm is presented. The synthetic speckle pattern serves as the calibration target. The particle swarm algorithm-based digital image correlation is employed to achieve matches among 3D control points and 2D image points; then the improved gray wolf optimizer algorithm is used to calculate the camera parameters. For verification, simulated and real tests are conducted. Through the analysis of calibration results, the proposed method performs better and is more stable than other calibration targets. Research on the influence of camera pose and optimization algorithm is conducted, showing that the improved gray wolf optimizer algorithm performs better than other benchmark algorithms. The camera parameters can be obtained through one captured image when the speckle patterns are added in the portion of the camera sensor.
相机校准对于各种基于视觉的三维计量技术至关重要。据我们所知，本文提出了一种新颖的相机校准方法，该方法结合了合成散斑图案和改进的灰狼优化算法。合成散斑图案用作校准目标。采用基于粒子群算法的数字图像相关技术来实现三维控制点与二维图像点之间的匹配；然后使用改进的灰狼优化算法来计算相机参数。为进行验证，开展了模拟测试和实际测试。通过对校准结果的分析，所提方法比其他校准目标表现更好且更稳定。对相机姿态和优化算法的影响进行了研究，结果表明改进的灰狼优化算法比其他基准算法表现更好。当在相机传感器部分添加散斑图案时，通过一张捕获图像即可获得相机参数。
