摘要
快速而准确地寻找到焊缝中心的位置是实时焊缝跟踪系统要解决的首要问题,文中提出一种基于视觉传感器CCD的焊缝跟踪系统。通过对焊缝图像的分析发现,焊缝图像具有较为明显的统计特征,即坡口的平均灰度较其它位置的平均灰度要大。基于这一统计特征,给出一种新的确定焊缝中心的快速算法。该算法根据像素点的灰度值对其进行排序,并以此为依据来确定焊缝边缘及焊缝中心。该算法无需进行滤波处理、对光源无特殊要求且鲁棒性强。现场实际计算表明所提算法准确可靠,适合于现场应用。
The problem of searching the center of weld fleetly and exactly is the key of automatic welding system, so a welding seam tracing system based on CCD was presented. By analyzing the gray values of welding seam image, the statistical properties on welding seam image were found that the average gray value of welding grooves is greater than the gray value of steel tube. Based on these statistical results, a new algorithm for detecting the seam center was put forward. The algorithm sequences all pixels according to each pixel' s gray value and the edge and core of seam are decided. The algorithm needs no filtration, specific lamp house and has good robustness. Examples show that the algorithm is nicety and practical.
出处
《焊接》
北大核心
2007年第4期53-55,共3页
Welding & Joining
基金
陕西省教育厅专项科研计划项目(06jk208
05jk218*)
关键词
埋弧焊
焊缝跟踪
图像处理
submerged arc welding, wdd tracing, image processing
