基于图像处理的焊管焊缝气孔检测

A new algorithm for detecting air holes in steel pipe welding based on image processing

提出对x射线焊缝图像进行图像处理以判断焊缝质量,首先利用Sobel算子结合二值化过程以确定焊缝位置。在明确了焊缝位置后,通过实验发现气孔处图像的灰度值较焊缝的灰度值至少高出5,根据这一气孔的视觉效果,提出一种快速图像处理算法以检测焊缝内是否存在气孔。该算法通过计算焊缝内有无灰度值高出周围像素点灰度值至少5的区域来判断气孔,因此计算速度快,适合于现场实际应用。实际计算表明,所提算法可行、有效。

We detect the quality of welding by an image-processing algorithm. In the presented algorithm, Sobel operator combined with image binarization is first used to decide the edge of welding. After many experiments, it is found that the gray value in the position of air hole is larger than the gray value in the position of welding at least 5. So, by examining whether there exists a region whose gray value is larger than the gray value of other parts of welding at least 5, the algorithm can detect air hole in the welding. The calculation speed of the algorithm is fast for it only processes the image in the welding. As an application, we successfully detect some air holes in real industrial welding x-ray images.