不等厚板激光焊接焊缝缺陷结构光视觉检测  

Defect detection of laser welding seam of unequal-thickness blank based on structured light vision

作　　者：高世一[1] 杨永强[1,2] 杨凯珍[1] 

机构地区：[1]广州有色金属研究院,广州510650 [2]华南理工大学机械与汽车工程学院,广州510641

出　　处：《激光技术》2011年第4期440-443,共4页Laser Technology

摘　　要：为了检测不等厚钢板激光焊接焊缝表面缺陷,采用结构光主动视觉检测法和高斯数据拟合技术进行了焊缝表面缺陷检测的实验研究。首先采用高斯拟合法对激光条纹图像进行处理,提取出具有亚像素精度的激光条纹中心线;然后通过最小二乘法拟合出3条相交直线并求其交点,近而获得精确的焊缝端点位置坐标;最后给出焊缝表面缺陷-凹度和凸度的检测方法,并以不等厚钢板激光焊接焊缝为检测对象进行验证。结果表明,这种基于拟合技术的结构光视觉检测法为判断激光焊接焊缝质量是否合格提供了较为准确的判断依据。In order to detect the laser welding seam surface defects of unequal-thickness steel blank, the structured light active vision method and Gaussian data fitting technology were used for the experiment research of the seam surface defects detection. Firstly, the laser strip center line in sub-pixel accuracy was extracted by Gaussian fitting. Secondly, the seam endpoint positions were obtained by line fitting and computing point of intersections of three lines. Thirdly, the computing methods of the surface defects of the laser welding seam, such as concavity, convexity, were put forward. The experiment result shows that the proposed method is based on structured light vision and data fitting is suit for laser welding seam detection of unequal-thickness steel blanks.

关 键 词：图像处理 焊缝表面检测 结构光视觉 高斯拟合 

分 类 号：TG456.7[金属学及工艺—焊接]