作　　者： ; ; ;

机构地区： 华南理工大学机械与汽车工程学院

出　　处： 《焊接学报》 2005年第6期25-27,34,共4页

摘　　要： 由于外部干扰小,用CCD采集的TIG焊熔池区图像很清晰,因此可以对该图像进行二值化处理。进而采用二值形态学方法分析了膨胀、腐蚀和开、闭四种基本运算的图像处理效果,并在此基础上从图像的预处理到边缘检测,做了更加深入的研究。由此,设计出了一套具有阈值自适应功能的二值形态学算法。试验结果表明,采用二值形态学对TIG焊视觉图像进行图像处理和特征提取,可以取得很好的效果,对后续的焊接质量控制和焊缝跟踪工作可以提供非常有用的熔池和焊缝信息。 Since the noise outside the molten pool of tungsten inert-gas (TIG) welding is very small, we change the image of molten-pool and seam to a binary image, and analyze the image processing results of the four fundamental binary morphological operations, including erosion, dilation, open and close. Based on these, we also work out a set of binary morphological operations which have the function of threshold self-adapting. From image pretreatment to edge detection, much more studies have be done. The results show that it is very effective to detect and process the charge coupled device image of tungsten inert-gas welding using binary morphology, which would offer very useful communications in welding process tracking and welding quality control.

关 键 词： 二值形态学 图像处理 焊 边缘检测

领　　域： [金属学及工艺]