高级检索
石端虎, 刚铁, 赵洪枫, 杨峰, 何敏. 双面T形焊件中单侧投影重合缺陷的自动判别方法[J]. 焊接学报, 2019, 40(4): 46-49. DOI: 10.12073/j.hjxb.2019400098
引用本文: 石端虎, 刚铁, 赵洪枫, 杨峰, 何敏. 双面T形焊件中单侧投影重合缺陷的自动判别方法[J]. 焊接学报, 2019, 40(4): 46-49. DOI: 10.12073/j.hjxb.2019400098
SHI Duanhu, GANG Tie, ZHAO Hongfeng, YANG Feng, HE Min. Automatic discrimination method of unilateral projection overlap defects in double T style weldments[J]. TRANSACTIONS OF THE CHINA WELDING INSTITUTION, 2019, 40(4): 46-49. DOI: 10.12073/j.hjxb.2019400098
Citation: SHI Duanhu, GANG Tie, ZHAO Hongfeng, YANG Feng, HE Min. Automatic discrimination method of unilateral projection overlap defects in double T style weldments[J]. TRANSACTIONS OF THE CHINA WELDING INSTITUTION, 2019, 40(4): 46-49. DOI: 10.12073/j.hjxb.2019400098

双面T形焊件中单侧投影重合缺陷的自动判别方法

Automatic discrimination method of unilateral projection overlap defects in double T style weldments

  • 摘要: 为了实现双面T形焊件中单侧投影重合缺陷空间位置数据的自动提取,在原有自动判别准则的基础上提出了单侧投影重合缺陷的自动判别方法. 该方法首先对焊件左转或右转射线检测图像进行逐行搜索,通过采用缺陷中心点距图像上端距离相同的方法找出单侧投影重合的缺陷,然后通过对比单侧投影重合缺陷的左右投影距离,判别重合缺陷在焊件内部的深度位置,从而将重合缺陷区分出来. 采用提出的方法对实际焊件中的单侧投影重合缺陷进行了自动判别,试验结果表明提出的方法是可行的.

     

    Abstract: In order to achieve the automatic extraction of spatial location data for unilateral projection overlap defects in double T style weldments, the automatic discrimination method of unilateral projection overlap defects was put forward on the basis of original automatic discriminant criterion. Firstly, line by line search was carried out towards X-ray detection images of weldments left turning or right turning, and the unilateral projection overlap defects were found by using the method of the same distance from defect center point to the upper end of the image, then the depth location of overlap defects were judged by comparing the left or right projection distance of unilateral projection overlap defects, and the overlap defects were discriminated. The automatic discrimination of unilateral projection overlap defects in actual weldments was carried out, and the experimental results show that the provided method is feasible.

     

/

返回文章
返回