Research Weld Defects Detection And Recognition Technology By X-ray Image

With the rapid development of the pipeline transportation industry,the safety of pipeline transportation has increasingly become the focus of attention.The quality of the pipe weld quality directly affects the safety of the pipe transportation,to ensure the quality of the pipe welds,it is necessary to detect the flaws in the welds.X-ray inspection has always been the main method of weld inspection because of its advantages such as high imaging speed and low cost.However,due to the influence of imaging methods and other factors,there are problems such as low contrast and high noise in the X-ray seam image,which makes the current defect detection algorithm less accurate.In actual production,X-ray weld images are mainly detected artificial,but this method is inefficient and subjective.Therefore,this paper focuses on the defect detection and identification of X-ray weld images.To solve the problems of low contrast in X-ray image,by analyzing the gray histogram of the weld seam image,we proposed an image enhancement algorithm based on adaptive gray stretching interval.We designed a self-Adapt the strategy to calculate the value range of the grayscale stretch interval.Experimental results show that this algorithm can effectively improve the contrast of the image compared with other algorithms,and achieved a more ideal result.There are many background areas in the X-ray weld image that are not related to the weld,and weld defects exist only in the weld area.By extracting the weld zone,the defect detection is limited to the weld zone,which greatly reduces the time required for detection and improves the accuracy of defect detection.Due to the influence of factors such as the pipe welding process and X-ray imaging technology,the welding seam has the characteristics of double edges and blurring edges,which causes great difficulties in the extraction of the weld zone.by analyzing several typical image segmentation algorithms,this paper proposes a weld seam extraction algorithm based on adaptive threshold and designs an adaptive threshold strategy to extract the weld seam edge.Experimental results show that the algorithm can effectively extract the weld area in the X-ray weld image.Defect detection is a very important part in the automatic detection of welds,and defect identification is an important basis for evaluating the quality of welds.For defectinspection,due to the low contrast between the weld defect and the weld seam and the vague edges of the defects,it is difficult to detect defects.By analyzing the characteristics of weld defects,we propose a weld defect detection algorithm based on grayscale morphology.The algorithm uses the open and closed operations in grayscale morphology to fit the background image of the weld,and through the background subtraction method to detect weld defects.The experimental results show that the algorithm can accurately extract the weld defects.For defect identification,by analyzing the defects of various welds and selecting a reasonable characteristic parameter,we train multiple classification models to classify defects and evaluate the merits of each model.