基于视觉传感的焊缝中心识别技术研究  

Study of technology for seam center detection based on visual sensor

作　　者：赵传敏[1] 高向东[1] 谢子方[1] 丁度坤[1] 

机构地区：[1]广东工业大学机电工程学院,广东广州510006

出　　处：《制造业自动化》2007年第8期45-47,57,共4页Manufacturing Automation

基　　金：国家自然科学基金资助项目(60375012);广东省自然科学基金资助项目(020176;06021444)

摘　　要：在焊接过程中通过CCD视觉传感器获取熔池和焊缝图像,采用多种滤波去噪技术和边缘检测技术对图像进行滤波和边缘检测试验,并对各种处理方法进行了比较和分析,得出了一种相对最佳的实时识别焊缝特征的组合图像处理方法。试验结果表明,对焊接图像采用Wiener滤波和Canny边缘检测能较好的提取焊缝中心位置的特征信息。The experiments were performed to study the technology for eliminating noises and detecting the seam center from images captured by a CCD visual sensor during the welding process. First, some filter methods were combined with the edge-detecting methods. Then, the best combinational method was found out to extract the seam characters from images by comparing the results of different methods. Finally, the analysis of the experiment results showed that the combination of Wiener filter with Canny edge detection was the most efficient to obtain the character information of the seam center.

关 键 词：视觉传感 图像处理 滤波 边缘检测 特征提取 

分 类 号：TG456.2[金属学及工艺—焊接]