基于小波边缘检测的焊缝跟踪技术  

Welding Seam Tracking Technology Based on Wavelet Edge Detection

作　　者：王宏文[1] 李松林[1] 曹思源[1] 黄巍巍[1] 

机构地区：[1]河北工业大学控制科学与工程学院,天津300130

出　　处：《热加工工艺》2012年第23期141-144,147,共5页Hot Working Technology

基　　金：河北省科学技术研究与发展计划项目(06212174D)

摘　　要：在基于视觉传感的焊缝跟踪系统中,焊缝图像复杂多变,要准确得到焊缝中心线,选取合理高效的处理算法是十分重要的。在详细分析焊缝图像特点基础上,采用先行基于梯度信息的自适应平滑滤波,运用3阶B样条双小波多尺度边缘检测方法,减少了误检漏检率,有效地提高了小波变换在抗噪性能和定位精度上的优越性。在此基础上提取出焊缝中心线,用于焊缝自动跟踪控制。试验结果表明,该方法复杂度适中,处理速度较快,效果良好,能有效地提高焊缝跟踪控制的精度。Based on vision sensing seam tracking systems, welding seam images are complex and changeable. It is very important to choose a reasonably efficient algorithm to accurately get the weld centerline. Based on analysing the characteristics of welding image, the adaptive filtering was used based on gradient information and double 3-order B-spline Wavelet edge detection method to reduce the false leak detection rate and effectively improve the performance and noise based on wavelet transform on the superiority of positioning accuracy. Then the weld centerline which is used on welding seam tracking control was extracted. Testing results show that this approach is moderately complexity, greatly effective and has faster processing and can effectively improve the precision of seam tracking control.

关 键 词：小波变换 多尺度边缘检测 误检漏检 焊缝跟踪 

分 类 号：TG441.7[金属学及工艺—焊接] TP391.41[自动化与计算机技术—计算机应用技术]