李冰 杨旭 何煊 翟永杰

摘  要： 為了提高薄板对焊焊接过程中焊缝识别的精度，基于斜率检测算法并结合斜率置信算法和斜率适应值，提出一种自适应薄板焊缝识别算法。通过对激光束在焊接工件上的三原色图像比较进行色位图选择，利用最大类间方差法提取图像中的激光束轮廓，对激光束轮廓按列计算质心行位置并进行线性拟合优化质心线数据，最后利用自适应斜率检测算法判断焊缝与焊板间的图像激光跳变位置，实现焊缝左右边界检测进而实现焊缝识别。试验结果表明，自适应焊缝识别算法可以实现对焊焊缝的检测与识别，与斜率检测算法相比，识别精度更高，有效地提高了焊接质量与焊接效率。

关键词： 焊缝识别; 焊接系统; 激光束轮廓; 质心线数据; 自适应斜率检测; 实验测试

中图分类号： TN911?34; TG409                  文献标识码： A                      文章编号： 1004?373X（2020）18?0182?05

Abstract： An adaptive thin plate welding seam identification algorithm is proposed based on the slope detection algorithm and combined with the slope confidence algorithm and the slope adaptation value to improve the welding seam identification accuracy during the thin plate butt welding process. The color bitmap is selected by comparing the three?primary color images of the laser beam on the weld workpiece， and the laser beam profile in the image is extracted by means of the maximum inter?class variance method. The centroid row position is calculated according to the column of the laser beam profile， and the centroid line data is optimized by means of the linear fitting. The image laser jumping position between the welding seam and the welding plate is determined by means of the adaptive slope detection algorithm， so that he left and right boundary detection of welding seam is realized， and then the welding seam recognition is realized. The experimental results show that the adaptive welding seam recognition algorithm can realize the detection and identification of the welding seam， in comparison with the slope detection algorithm， the recognition accuracy is higher， and the welding quality and welding efficiency are effectively improved.

Keywords： welding seam identification; welding system; laser beam profile; centroid line data; adaptive slope detection; experimental measurement

0  引  言

由于焊接过程中产生的强光、烟尘等会危害到人体健康，同时人工焊接存在焊接质量不稳定、效率低等问题，因此传统的人工焊接方式已逐渐被自动焊接所替代[1]。自动焊接中进行焊缝检测的传感系统[2]主要有电弧传感系统和视觉传感系统2种。随着电荷耦合元件（Charge?Coupled Device， CCD）等视觉传感器以及数字图像处理算法等领域的发展，视觉传感系统以其信息广泛、灵活智能等优点在焊接控制和焊缝跟踪等领域得到了广泛应用[3?4]。其中，如何实现实时精确地焊缝位置跟踪成为视觉传感系统研究的热点与难点[5?6]。……