分形理论和LOWESS方法在球罐焊接坡口检测中的应用

针对球罐焊接过程坡口边缘灰度变化不明显 ,受噪声和其它干扰信息的影响很难准确及时地处理图像的弱点 ,本文提出了分形理论和LOWESS方法相结合的图像处理方法。该方法是根据坡口图像的特点 ,首先利用LOWESS方法通过拟合坡口图像每一行的灰度值最低点求出坡口的中心线 ,然后利用图像的分形特征求坡口边缘相对于坡口中心线的偏移量 ,从而进行坡口边缘的识别。实验结果证明 ,该方法不仅能准确地识别出坡口图像的边缘 ,而且节省了图像处理的时间 ,为焊接过程的实时控制提供了可靠的信息

Application of Fractal Theory and LOWESS Method in Detecting Spherical Tank Welding Groove

It is difficult to detect the edges of spherical tank welding groove because the edges showed no significant changes in gray value, and the traditional method of image processing is sensitive to noise and other disturbances. To overcome these difficulties, a new method is proposed in this paper which is the combination of fractal theory and LOWESS method. Based on the characters of the groove image, central line of the seam was fitted with the minimum gray value of each line using the LOWESS method and offset between the edges and central line obtained according to fractal character of the image, then image edges were recognized. Experiment results show that the image edge of the groove can be recognized correctly and quickly. The new developed image processing method could provide reliable information for the real time control of welding process.