摘要
为了构建一种基于单摄像机的三维视觉系统,提出了重构光学标定点三维坐标、计算重构误差和优化标定参数的方法。通过移动高精度位移平台,构建三维视觉系统,计算左右摄像机位的初始参数,引入质心坐标法计算两摄像机位间的旋转平移矩阵,利用最优三角剖分法重构光学标定点的三维坐标,计算并最小化重构误差,对标定参数进行优化。实验表明:重构误差直接反映了三维重建的效率,该方法的精度和鲁棒性得到了显著的提高;实物外形测量的误差从0.1123 mm减小到0.0191 mm,标准差从0.1838减小到0.1275。该方法适合应用于三维视觉系统标定质量的评估。
A new method is proposed to optimize the calibration parameters of the parallel binocular vision system based on single camera by calculating the reconstruction error of the three-dimensional coordinates of the standard points.Using a single camera and a high-precision displacement platform to realize binocular vision system,the initial parameters of the left and right cameras were calculated.The centroid coordinates were introduced to calculate the rotation translation matrix between the two cameras,and the three-dimensional coordinates of the standard points were reconstructed by triangulation.The reconstruction errors were calculated and minimized,and the calibration parameters were optimized.The reconstruction error directly reflects the efficiency of three-dimensional reconstruction and is more suitable for the evaluation of three-dimensional vision system calibration quality.The test results show that the accuracy and robustness of this method are improved significantly.The physical shape measurement error and standard deviation are reduced from 0.1123 mm to 0.0191 mm and from 0.1838 to 0.1275.
作者
褚鑫磊
隋国荣
CHU Xinlei;SUI Guorong(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)
出处
《光学仪器》
2021年第5期16-24,共9页
Optical Instruments
关键词
单摄相机
光学标定
三维重建
质心坐标法
三角剖分
single camera
optics calibration
three-dimensional reconstruction
centroid coordinate method
triangulation
