中文摘要：

对一种新型高效多弧焊接方法——旁路耦合电弧MIG焊（DE-GMAW）在连续大电流焊接条件下的焊接熔池图像进行被动视觉采集和熔池边缘提取算法研究.为降低熔池周围电弧信号对边缘提取结果的影响,提出一种新的图像预处理算法,并在对比6种微分算子边缘提取结果的基础上,提出一种新的熔池边缘提取算法.结果表明：新的图像预处理算法可以有效降低边缘提取时熔池周围电弧信号的干扰;新的边缘提取算法在连续大电流焊接条件下,可以得到准确度、连续性和平滑度均符合要求的旁路耦合电弧MIG焊焊接熔池边缘信息.

英文摘要：

Double-electrode MIG welding（DE-GMAW） is a new kind of high-efficiency MIG welding.The acquisition of passive vision was used for the weld pool image of DE-GMAW with continuous and large weld current.And the edge extraction algorithm of the weld pool image was studied.In order to reduce the effect of weld arc signal around the weld pool on its edge extraction,edge extraction with a new kind of image preprocessing algorithm had been proposed.And based on the comparison result of edge extraction with 6 differential operators,a new kind of edge extraction algorithm of weld pool image was proposed,also.The result showed that the new method of image preprocessing was able to reduce effectively the interference of the arc signal around the weld pool in the edge extraction.And the new method of edge extraction algorithm could be used to extract the edge message of weld pool with an accuracy,continuity and smoothness under the condition of continuous high weld current,so that the requirement of the DE-GMAW was met.