We investigate the light strip center extraction and system calibration method of a robot line structured light vision measurement system to realize fast on-line measurement of complex objects. First, the model of a robot vision measurement system is established to analyze the key technologies involved for improving the measurement accuracy of the system. Second, on the basis of the traditional gray gravity method, a gray gravity method based on the normal direction is proposed to extract the center of the laser stripe. Considering the influence of camera distortion, the accuracy of extracting the center stripe can be further improved. Third, a two-dimensional checkerboard is applied to synchronously calibrate the camera parameters, laser plane parameters, and hand-eye relationship of the robot, thus improving the efficiency and accuracy of system calibration. Finally, three-dimensional topography and the geometric dimensions of a complex part are measured by the constructed system with high accuracy. The experimental results show that the average measurement error is <0.2 mm and the measurement accuracy is <1.3 % , indicating that the system can be effectively used in actual industrial measurement.

中文翻译：

---

机器人线结构光视觉测量系统：灯带中心提取及系统标定

我们研究了机器人线结构光视觉测量系统的光带中心提取和系统校准方法，以实现复杂物体的快速在线测量。首先，建立机器人视觉测量系统模型，分析提高系统测量精度所涉及的关键技术。其次，在传统灰度重力法的基础上，提出了一种基于法线方向的灰度重力法提取激光条纹中心。考虑到相机畸变的影响，可以进一步提高提取中心条纹的精度。第三，采用二维棋盘格同步标定机器人的相机参数、激光平面参数、手眼关系，从而提高系统校准的效率和准确性。最后，构建的系统可以高精度地测量复杂零件的三维形貌和几何尺寸。实验结果表明，平均测量误差<0.2 mm，测量精度<1.3%，表明该系统可有效应用于实际工业测量。