焊缝视觉跟踪系统中标定算法的研究

Study of Calibration in Auto-vision Welding Seam System

针对TIG焊焊缝视觉跟踪系统的标定问题,基于光学成像的几何原理,通过CCD摄像 机获取焊接区熔池图像,经图像处理后,将其转化成图像目标坐标矩阵,以转换后的目标矩阵为初 始矩阵,确立了一种可以完成4种坐标相互转换的数学模型,以此数学模型可以完成三雏世界坐标 与二维图像坐标之间的相互转换.将研究结果应用于铝合金TIG焊过程的质量控制,保证了焊接过 程的稳定,实现了对铝合金TIG焊焊缝成型的良好控制.

A method of calibration is developed in terms of accuracy and stabilization of auto -vision TIG welding. Welding image is captured by CCD camera. After image processing, the image is transformed by coordinate matrix, then conversion between world coordinate, camera coordinate, image coordinate and pixel coordinate are carried out according to the principle of photic geometry. The mathematic model realizes conversion between three -dimension world coordinate and two - dimension image coordinate. Created model is mainly applied to quality control of TIG aluminum alloy welding. The model ensures the stabilization of welding process and a successful control of the formation of TIG aluminum alloy welding.