油管螺纹图像的Canny算子边缘提取方法研究

石油生产中油管螺纹失效的问题相当突出，采用图像处理的方法取代传统方法使油管螺纹检测效果得到极大改善。根据油管螺纹图像的特点，针对传统的Canny算子中高斯平滑滤波的缺陷，采用中心加权MTM滤波算法，并且将像素的八邻域分为4个象限，结合双线性插值的方法取代非极值抑制方法，进一步提高边缘检测的精度。试验表明，相对于LOG算子，改进后的Canny算子更好地提高了油管螺纹的检测精度。

Study of Edge Detection Based on Canny Operator in Tubing Threads Images Processing

The invalidation of tubing threads is serious in normal oil production. A great improvement on tubing threads measuring was made to substitute image processing for manual methods. According to the characteristic of tubing threads＇ images and limitation of Gauss filter, an improved edge detection algorithm was used, which replaced Gauss filter with central coefficient MTM filter and a four-quadrant bilinear interpolation was applied to the eight neighborhoods of pixel gradients based on Canny operator to improve the precision of the detected edge. Experimental results show that the new Canny operator improved the precision in measuring tubing threads compared with LOG operator.