Based on binocular vision, this paper proposes an algorithm for fast and accurate three-dimensional measurement. The binocular 3D measurement system is mainly divided into four parts: binocular calibration, binocular correction, binocular matching and pose calculation. In this paper, an international softwareM is used to achieve camera calibration, the FAST algorithm is used to achieve the extraction of feature points, and the matching method that combines gray-level correlation coefficient (cc) and epipolar constraint is selected. Experiments have verified that this algorithm can accurately achieve three-dimensional pose measurement.

中文翻译：

---

基于双目视觉的3D位姿测量算法研究

本文基于双目视觉，提出了一种快速准确的三维测量算法。双目3D测量系统主要分为双目标定、双目校正、双目匹配和位姿计算四个部分。本文采用国际软件M实现相机标定，采用FAST算法实现特征点提取，选择灰度相关系数（cc）和对极约束相结合的匹配方法。实验证明，该算法能够准确实现三维位姿测量。