基于亚像素的埋弧焊焊缝图像处理

摘要

Submerged arc welding is one of the widespread welding technology.It has high efficiency,good weld quality.no arc radiation, etc.It is more and more important/To improve the submerged arc welding process automatic and intelligent, increasing demands for precision tracking technology.In order to improve the accuracy of submerged arc welding, a sub-pixel location based on B Spline transform was put forward for welding line and dibided edge.The algorithm is conducting a Marr-Hildreth edge detection and a sub-pixel location.Field tests,achieved standard error less than 255.The traditional edge detection was low in precision,but the problem is constraining by using the transform.%埋弧焊是普遍使用的熔焊方法之一,具有生产效率高、焊缝成形好、没有弧光辐射等诸多优点,应用前景广阔.为了提高埋弧焊焊接过程的自动化和智能化,对埋弧焊精密跟踪技术的要求不断提高,需对焊缝间隙和坡口边缘进行精密检测.在此提出了一种基于B样条变换的亚像素定位算法,该算法是在对图像进行一次Marr-Hildreth边缘检测后再进行亚像素二次提取.通过现场试验,实现了检测标准误差小于255E/μm,克服了传统边缘检测精度低的问题.