立体视觉测量系统标定误差补偿

双目视觉测量系统利用三角测量法原理进行三维测量，其结构特性决定了测量误差随测量距离的增加而增大。针对测量误差的分布规律，该文提出基于局部视场的双目视觉测量系统优化方法。利用外在结构建立测量坐标系，减小标定和测量过程信息非一致性引入的系统误差；利用相机参数之间的耦合作用补偿系统固有参数标定误差并建立查找表，建立关于标定参数的虚拟映射。模拟实验最大误差小于0.03 mm，系统实验误差小于0.3%，实验表明优化后系统主要测量误差来源于探测器离散化引入的随机误差，双目视觉测量系统达到理论测量精度。

Calibration error compensation technique of stereoscopic vision measurement system

The principle of the triangulation is used to conduct the 3D measurement by binocular vision measurement system, and its structural characteristics determine that the measurement error increases with the increase of the measurement distance. Aiming at the distribution law of the measurement error, the optimization method of the binocular vision measurement system based on the local FOV was proposed. The measurement coordinate system was established by using the external structure to reduce the systematic errors introduced by the inconsistence of the calibration and measurement process information. The coupling effect between camera parameters was used to compensate the calibrated error of the system inherent parameter, and a look-up table was established to realize the virtual mapping of the calibration parameters. The results show that the maximum error is less than 0.03 mm in simulation experiment, and the measurement error is less than 0.3% in systematic experiment. After the optimization, the main measurement error comes from the random error introduced by the detector discretization, and the binocular vision measurement system achieves the theoretical measurement accuracy.