基于OpenCV处理视频图像检测焊缝的研究  

Research about weld detection based on OpenCV dealing with video image

作　　者：汪欣[1] 范立南[1] 张广渊[2] 陈切春[3] 

机构地区：[1]沈阳大学信息工程学院,沈阳110044 [2]山东交通学院,济南250023 [3]日本东北大学工学研究科

出　　处：《仪器仪表用户》2012年第3期1-4,共4页Instrumentation

基　　金：辽宁省自然科学基金(20102154);辽宁省教育厅科研项目计划(L2010376)

摘　　要：提出了一种基于处理视频图像对焊缝进行检测的方法。将视频文件中的原始焊缝图像作为研究目标,采用灰度变换法对原始图像进行预处理,选用Canny算子对焊缝图像进行边缘检测,并运用改进的Hough变换进行线段检测,得到焊缝坡口的中心线以及中心线的中点位置,实现了对焊缝焊点的检测。实验证明,该方法实现成本低,实时性强,精确度较高,智能性强。Putting forward a weld detection method which based on dealing with video image. In this methed, uses the orig- inal weld image which in video files as the research object, uses the gray transform to preprocess the original image, se- lect the Canny operator to detect the weld image edge, and use the Hough transform algorithm for line detection, get the weld groove centerline location,and the location of the centerline and the centerline. Realized the detection of weld and Solder joints. The experimental results proved that this method has some advantages of low cost, good real-time, high precision, strong intelligence.

关 键 词：灰度变换 CANNY算子 焊缝边缘检测 HOUGH变换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]