Visual sensor plays an important part in intelligentized welding systems, and the calibration of the vision sensor is the indispensable part of visual systems. Aiming at the problem of the tedious calibration process, this paper describes an automatic calibration algorithm. First, the robot motion equation and the motion range constraint equation are proposed to ensure that the collected images of calibration grid and laser line meet the calibration requirements. Based on these two equations, the automatic collection procedure can be realized. Second, the simplified visual servoing method and the Extended Kalman filter were used to adjust the images and rectify system parameters, respectively, which will improve the stability of the calibration motion. Third, to reduce the impact of the complex welding environments, a robustness feature extraction algorithm based on local threshold is studied. And then, the laser plane and hand-eye matrix are fitted with optimization algorithms to ensure calibration accuracy. Finally, the simulation experiments prove the feasibility and stability of the proposed algorithm. And the actual calibration tests suggest that the algorithm can significantly improve calibration efficiency. Moreover, the experimental results of welding guidance and seam tracking confirm that the calibration precision has met the welding requirement.

视觉传感器在智能化焊接系统中起着重要的作用，视觉传感器的校准是视觉系统必不可少的部分。针对标定过程繁琐的问题，提出了一种自动标定算法。首先，提出了机器人运动方程和运动范围约束方程，以确保所收集的校准网格和激光线图像满足校准要求。基于这两个方程，可以实现自动收集程序。其次，分别使用简化的视觉伺服方法和扩展卡尔曼滤波器来调整图像并校正系统参数，这将提高校准运动的稳定性。第三，为减少复杂焊接环境的影响，研究了基于局部阈值的鲁棒性特征提取算法。然后，在激光平面和手眼矩阵上安装优化算法以确保校准精度。最后，仿真实验证明了该算法的可行性和稳定性。实际的校准测试表明该算法可以显着提高校准效率。此外，焊接指导和焊缝跟踪的实验结果证明校准精度已满足焊接要求。实际的校准测试表明该算法可以显着提高校准效率。此外，焊接指导和焊缝跟踪的实验结果证明校准精度已满足焊接要求。实际的校准测试表明该算法可以显着提高校准效率。此外，焊接指导和焊缝跟踪的实验结果证明校准精度已满足焊接要求。