基于连通域分析的钢管焊缝缺陷检测方法

Defect Detection Method for Steel Pipe Welds Based on Connected Domain Analysis

中文关键词:  焊缝缺陷  图像处理  边缘检测  Hough变换  OTSU算法

英文关键词:weld defect  image processing  edge detection  Hough transformation  OTSU algorithm

中文摘要:

      焊缝缺陷检测是工业环节中极其重要的一环，焊接的质量直接影响工程的质量，因此焊缝缺陷的检测变得尤为重要。针对传统人工检测存在的效率低、误判率高、细微缺陷检测准确率低、工作强度大等不足，提出了一种基于连通域分析的钢管焊缝缺陷检测方法。首先，利用X射线成像系统得到焊缝原始图像；然后，经过Canny边缘检测、Hough变换和条件筛选提取出完整的焊缝区域；之后，通过OTSU算法、梯度法和连通域分析，填充了缺陷内部以获取完整的焊缝缺陷；最后，计算缺陷位置的平均灰度并根据筛选条件选择出真正的焊缝缺陷，自动检测并标记了焊缝原始图像中的缺陷，有效降低了焊缝缺陷检测的漏检率和误报率。实验测试了1 000幅左右的焊缝原始图像，验证了所提算法的正确性和有效性。所提算法识别率可达95.3%，漏检率为2.1%，误报率为2.6%。对比其他算法，所提算法对缺陷类型不敏感，检测速度快且识别率高，具有较强的适应性和通用性。

英文摘要:

      Weld defect detection is an extremely important part of the industrial link,and the quality of welding directly affects the quality of the project,so the weld defect detection is very important.In view of the defects of traditional manual inspection,such as low efficiency,high misjudgment rate,low accuracy of subtle defect detection and high work intensity,a defect detection method for steel pipe welds based on connected domain analysis is proposed.Firstly,X-ray imaging system is used to obtain the original image of the weld.Then,the complete weld area is extracted by Canny edge detection,Hough transform and screening conditions.After that,the complete weld defects are obtained by OTSU algorithm,gradient method and connected domain analysis.Finally,the average gray level of the defect location is calculated and the real weld defects are selected according to the screening conditions,and the defects in the original image of the weld are automatically detected and labeled,which effectively reduces the missed detection rate and false positive rate of the weld defect detection.The correctness and effectiveness of the proposed algorithm are verified by testing about 1 000 original images of welds.The recognition rate of this algorithm is 95.3%,the missed detection rate is 2.1%,and the false positive rate is 2.6%.Compared with other algorithms,the proposed algorithm is insensitive to defect types,has fast detection speed and high recognition rate,and has strong adaptability and universality.