基于数学形态学的熔池图像处理  

Image processing of molten pool based on mathematical morphology

作　　者：刘明涛[1] 高向东[1] 陈建辉[1] 谢子方[1] 

机构地区：[1]广东工业大学机电工程学院,广州市510006

出　　处：《焊接》2008年第5期29-32,共4页Welding & Joining

基　　金：国家自然科学基金(60375012);广东省自然科学基金(6021444;07001764);广东省科技计划项目(2007B010400069)

摘　　要：建立了用于气体保护钨极电弧焊(GTAW)的CCD视觉传感系统。在一定的工艺条件下,通过视觉传感器采集较为清晰的焊接熔池图像。选择包括熔池的区域为处理区域,对该区域进行中值滤波,用数学形态学中的高-低帽变换增强熔池图像的对比度。然后对图像进行二值化处理,用基本数学形态学的腐蚀、膨胀、开启、闭合对图像进行处理,用边缘检测算子提取熔池边缘并细化,获得了比较好的效果。A CCD vision sensor system was built to be applied in GTAW.A group of molten pool images had been gotten in a certain technological conditions by the vision sensor.And the area including the whole pool was chosen.First,the zone was processed by the median filter,and the contrast of the molten pool was strengthened by using top-bot method of mathematical morphology.Then,the image was changed to binary image,the theories of mathematical morphology,including dilation,erosion,open and close,were used to deal with the image.Finally,morphology edge detection algorithm was used to detect the pool edge and make it clear.The conclusion proved that the method is effective.

关 键 词：数学形态学 熔池图像 高-低帽变换 边缘检测 

分 类 号：TG409[金属学及工艺—焊接]