强弧光干扰下焊缝结构光图像跟踪信息获取  

Extraction of structure light seam tracking information in the presence of strong arc light interference

作　　者：程柏[1] 郑军[1] 潘际銮[1] 

机构地区：[1]清华大学机械工程系,先进成形制造教育部重点实验室,北京100084

出　　处：《清华大学学报（自然科学版）》2008年第11期1900-1903,共4页Journal of Tsinghua University(Science and Technology)

基　　金：先进成形制造教育部重点实验室开放基金(2006011)

摘　　要：利用激光结构光进行焊缝跟踪时,在存在强弧光干扰的情况下稳定地获取跟踪信息是非常关键一个问题。采用模板匹配的方法实时地计算得到当前图像中的感兴趣区域,并且利用重心法结合改进的一维K a lm an滤波的方法准确的提取出强弧光干扰下感兴趣区域内激光结构光的中心线,最后利用中心线上各点的横坐标计算加权系数对纵坐标进行加权累积,稳定并且准确地获得了能够体现出整条结构光变化的跟踪点信息。实验表明,该文提出的算法具有很强的抗干扰能力,可以在实际的焊缝跟踪系统中得到应用。When structured laser light is used for seam tracking, strong arc light interference significantly complicates stable extraction of the tracking information. This paper analyzes the real-time region of interest using the template matching method, with the gravity method and the modified one-dimensional Kalman filtering method used to stably extract the center line in the presence of strong arc light interference. The abscissas of all the points on the center line are used to calculate the weight coefficients to obtain the weighted sum of the corresponding ordinate value to precisely extract the tracking information. Experiment show that the algorithm provides strong anti interference capabilities in practical seam tracking systems.

关 键 词：焊缝跟踪 模板匹配 改进Kalman滤波 加权算法 

分 类 号：TG409[金属学及工艺—焊接] TN911.73[电子电信—通信与信息系统]