基于霍夫变换的管子管板焊缝图像的缝隙检测算法

Gap Detection Algorithm of Tube and Tube Sheet Weld Image Based on Hoff Transformation

在管子管板焊缝图像检测中,探测器是由小成像板拼接而成,导致采集到的成像图像中存在缝隙。因此,利用成像图进行缺陷检测之前,需要将图像中的缝隙检测出来。为此,本文提出了一种基于霍夫变换的焊缝图像缝隙检测算法,研究了在直线缝隙检测前的预处理过程,包括利用中值滤波方法去除图像噪声,利用Canny边缘检测方法检测图像边缘。最后,着重研究了提取所需缝隙,剔除无效的成像板轮廓的算法。实验结果表明,该方法能正确的检测出直线缝隙区域,取得了比较理想的检测效果。

In the test of tube and tube sheet weld, the detector is made up of small image plates, which can cause a gap in the image. As a result, the gaps in the diagram need to be detected before using an imaging map for defect detection. For this reason, this paper proposes a weld image crack detection algorithm based on Hough Transform, studied in front of the straight line crack detection pretreatment process, including the use of median filte r in g method to remove the image noise, using Canny edge detection method to detect image edge. Finally, the algorithm for extracting the desired gap and eliminating invalid image plate contour is studied. The experimental results show that the method can detect the area of the straight gap and obtain the desired results.