Crack defect detection method used in industrial welding seam X-ray image

Abstract

The invention belongs to the technical field of industrial welding seam defect detection, and particularly relates to a crack defect detection method for an industrial welding seam X-ray image, and the method comprises the following steps: carrying out self-adaptive histogram equalization processing with low limit contrast on a to-be-detected industrial welding seam original X-ray image; performing de-noising processing on the equalized image by adopting non-local mean filtering to obtain a de-noised X-ray image; performing class edge texture detection on the de-noised X-ray image by adopting a Canny algorithm to obtain a preliminary detection result; and screening the preliminary results to obtain crack defect detection results. According to the method, the result obtained through the Canny algorithm is further filtered and screened, the edge of the welding seam can be removed to a great extent through the algorithm, only the crack type defects in the welding seam are reserved, and the detection rate of the crack type defects in the final result is extremely high.

Description

technical field [0001] The invention belongs to the technical field of industrial weld seam defect detection and the technical field of radiographic flaw detection, and in particular relates to a method for detecting crack defects in industrial weld seam X-ray images. Background technique [0002] In the context of today's industrial society, welding has become the most popular connection technology in the industrial field. No connection technology is so widely and commonly used in various fields like welding. Welding has become a key manufacturing technology. As one of the assembly processes, welding is usually arranged in the later or final stage of the manufacturing process, so it has a decisive effect on product quality. Because of this, welding is seen as a critical manufacturing technique in many industries. As a large industrial country, our country has a huge welding workload, and whether it is machine or manual welding, there will always be more or less defects, an...

AI Technical Summary

Problems solved by technology

However, this method is not effective in detecting cracks with long dimensions and irregular shapes.

[0006] From the above analysis, it can be seen that most of the current methods are still based on magnetic particle or ultrasonic detection, which is limited in technical operation, and the features contained in the images obtained by these methods cannot fully express structural information and lack features; the transformation of original features is relatively difficult. Many, the post-screening calculation method is complex

Unable to detect cracks with peculiar and complex shapes

Images