板焊缝图像处理  

The Image Management of Welding Line for Plates

作　　者：刘冬花[1,2] 马国红[1] 李琦[1] 

机构地区：[1]南昌大学机电工程学院,江西南昌330031 [2]江西技师学院,江西南昌330200

出　　处：《中国制造业信息化（学术版）》2011年第8期64-66,共3页

摘　　要：焊接过程中不可避免地受到弧光、飞溅以及烟尘等若干因素干扰,为了提取准确、清晰的焊缝图像,提出了利用中值滤波进行图像预处理,采用固定阈值的方法对图像进行分割,应用Canny边缘算子进行边缘检测,采用Sobel算子对边缘进一步细化,采用最小二乘法进行曲线拟合以得到更为精确的图像,最后采用几何中心法较为准确地提取焊缝中心。以上各过程都是利用Visual C++实现,实验结果表明,方法快速有效,原理明了,算法简单,易于编程。The welding process is inevitable and related to arc, splash, smoke and dust. In order to extract accurate and clear weld image, it describes median filtering image preprocess. This method uses the fixed threshold segmentation process of image edge operator, applies Canny edge detection and Sobel edge operator tO further refinement, emploies least square method in curve fitting to get more precise image. Based on above process, it develops the application system with Visual C ＋ ＋. The experiment result shows this method is efficient, simple and clear, the principle is simple and easy to programming.

关 键 词：板焊缝 图像处理 VISUAL C++ 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]