水下焊接机器人视觉传感系统图像干扰因素分析

Analyses on interference factors on image in underwater welding robot vision sensor system

水下焊接机器人视觉传感系统在焊接过程中实时的对焊缝图像进行提取,通过一系列的图像处理算法,识别出焊缝位置,得到焊缝与焊枪之间的位置偏差,通过控制算法,实现焊缝跟踪的目的。在图像提取的过程中,会有各种各样的干扰,如泥沙干扰、弧光干扰、水泡气泡干扰、水对光的散射等各种干扰。着重分析了水下焊接机器人视觉传感系统图像提取的各种干扰因素,并对各种干扰因素产生的原理加以研究分析,并在此基础上,对如何去除各种干扰因素的方法——"二步干扰因素去除法"进行了简要的分析。

The underwater welding robot vision sensor system extracts seam images in the process of welding.Through a series of image processing algorithms,seam position was identified,then the deviation between the seam and torch was obtained.At last,seam tracking was realized by control algorithms.In the process of extracting images,there will be various interference,such as silt interference,arc interference,blisters bubble interference,water light scattering interference and other interferences.All kinds of interference factors on the image extraction of underwater welding robot vision sensor system were analyzed,and the generation of all kinds of interference factors were studied.The method to remove all kinds of interference was proposed based on the above analyses,which is called'two-step removal of interference factors method'.