基于X射线探伤的焊接图像缺陷自动检测方法

传统的图像缺陷自动检测方法存在对缺陷识别遗漏的问题，为此引进X射线探伤技术，设计一种针对焊接图像的缺陷自动检测方法，通过动态叠加焊接图像多帧信息的方式，降低焊接图像内噪声，提高图像整体分辨率。同时，绘制焊接图像典型缺陷示意图，以此为参照，使用X射线滤波算法，自动提取帧内缺陷信息，并根据图像中不同缺陷信息的表达方式，对焊接图像缺陷进行动态化检测，以此完成对方法的设计。在此基础上，设置对比实验，证明相比对比的缺陷自动检测方法，设计的方法能够保证检测正确率达到98%。

Automatic detection method of welding image defects based on X-ray flaw detection

The traditional image defect automatic detection method has the problem of missing defect identification.Therefore,an automatic defect detection method for welding image is designed by introducing X-ray flaw detection technology.By dynamically superimposing multi frame information of welding image,the internal noise of welding image is reduced and the overall resolution of image is improved.At the same time,draw the typical defect diagram of the welding image,take this as a reference,use the X-ray filtering algorithm to automatically extract the defect information in the frame,and dynamically detect the defects in the welding image according to the expression of different defect information in the image,so as to complete the design of the method.On this basis,comparative experiments are set up to prove that compared with the comparative automatic defect detection method,the designed method can ensure the detection accuracy of 98%.